\makeatletter \@ifundefined{HCode} {\documentclass[screen, CRGEOS, Unicode, Thematic]{cedram} \renewcommand\bibsection{\section*{\refname}} \newenvironment{noXML}{}{} \def\thead{\noalign{\relax}\hline} \def\endthead{\noalign{\relax}\hline} \def\tbody{\noalign{\relax}\hline} \def\hyphen{\mbox{-}} \def\tsup#1{$^{{#1}}$} \def\tsub#1{$_{{#1}}$} \def\0{\phantom{0}} \def\mn{\phantom{$-$}} \def\botline{\\\hline} \usepackage{amssymb,wasysym} \usepackage[figuresright]{rotating} \RequirePackage{etoolbox} \def\jobid{crgeos20230202} %\graphicspath{{/tmp/\jobid_figs/web/}} \graphicspath{{./figures/}} \def\myRef{\ref} \newcounter{runlevel} \usepackage{multirow} \def\nrow#1{\@tempcnta #1\relax% \advance\@tempcnta by 1\relax% \xdef\lenrow{\the\@tempcnta}} \def\xmorerows#1#2{#2} \def\morerows#1#2{\nrow{#1}\multirow{\lenrow}{*}{#2}} \def\tabnote#1{\vskip4pt\parbox{.87\linewidth}{#1}} \def\xxtabnote#1{\vskip4pt\parbox{.7\linewidth}{#1}} \def\xxxtabnote#1{\vskip4pt\parbox{.98\linewidth}{#1}} \def\og{\guillemotleft} \def\fg{\guillemotright} \let\MakeYrStrItalic\relax \def\refinput#1{} \csdef{Seqnsplit}{\\} \def\ubreak{\newline} \def\back#1{} \def\para{\par} \def\vsptab{\vspace*{20pc}} \makeatletter \g@addto@macro{\UrlBreaks}{\UrlOrds} \gappto{\UrlBreaks}{\UrlOrds} \makeatother \DOI{10.5802/crgeos.226} \datereceived{2023-02-24} \daterevised{2023-06-22} \dateaccepted{2023-06-22} \ItHasTeXPublished }{ \PassOptionsToPackage{authoryear,comma}{natbib} \documentclass[crgeos]{article} \def\CDRdoi{10.5802/crgeos.226} \let\newline\break \let\ubreak\relax \def\xmorerows#1#2{\morerows{#1}{#2}} \makeatletter \def\CDRsupplementaryTwotypes#1#2{} \newenvironment{sidewaystable}{\begin{table*}}{\end{table*}} \def\xxtabnote{\tabnote} \def\xxxtabnote{\tabnote} \def\href#1#2{\url[#1]{#2}} \def\vsptab{} } \makeatother \def\0{\phantom{0}} \usepackage{upgreek} \usepackage[T1]{fontenc} \dateposted{2023-07-19} \begin{document} \begin{noXML} %\makeatletter %\def\TITREspecial{\relax} %\def\cdr@specialtitle@english{Magma degassing and its impact on the Earth's atmosphere: from magma oceans to lava lakes} %\def\cdr@specialtitle@french{Impact atmosph\'erique du d\'egazage magmatique : des oc\'eans de magma aux lacs de lave} %\makeatother \title{Influence of volatiles (H$_{2}$O and CO$_{2}$) on shoshonite phase equilibria} \shortrunauthors \author{\firstname{Francesco} \lastname{Vetere}\CDRorcid{0000-0002-0723-1990}\IsCorresp} \address{Department of Physical Sciences, Earth and Environment, University of Siena, 53100, Italy} \address{Leibniz University of Hannover, Institute for Mineralogy, Callinstrasse 3, Hannover, D-30167, Germany} \email[F. Vetere]{francesco.vetere@unisi.it} \author{\firstname{Olivier} \lastname{Namur}\CDRorcid{0000-0001-9129-3264}} \addressSameAs{2}{Leibniz University of Hannover, Institute for Mineralogy, Callinstrasse 3, Hannover, D-30167, Germany} \address{Department of Earth and Environmental Sciences, KU Leuven, 3001 Leuven, Belgium} \email[O. Namur]{olivier.namur@kuleuven.be} \author{\firstname{Francois} \lastname{Holtz}\CDRorcid{0000-0002-2043-8603}} \addressSameAs{2}{Leibniz University of Hannover, Institute for Mineralogy, Callinstrasse 3, Hannover, D-30167, Germany} \email[F. Holtz]{f.holtz@mineralogie.uni-hannover.de} \author{\firstname{Renat} \lastname{Almeev}\CDRorcid{0000-0003-0652-9469}} \addressSameAs{2}{Leibniz University of Hannover, Institute for Mineralogy, Callinstrasse 3, Hannover, D-30167, Germany} \email[R. Almeev]{r.almeev@mineralogie.uni-hannover.de} \author{\firstname{Paola} \lastname{Donato}\CDRorcid{0000-0002-4287-7543}} \address{Department of Biology, Ecology and Earth Sciences, University of Calabria, 87036 Arcavacata di Rende CS, Italy} \email[P. Donato]{paola.donato@unical.it} \author{\firstname{Francesco} \lastname{Frondini}\CDRorcid{0000-0002-7539-9541}} \address{Department of Physics and Geology University of Perugia, piazza Universit\`{a}, 06100 Perugia, Italy} \email[F. Frondini]{francesco.frondini@unipg.it} \author{\firstname{Michele} \lastname{Cassetta}\CDRorcid{0000-0003-2078-655X}} \address{Department of Computer Sciences, University of Verona, I-37134 Verona, Italy} \address{Department of Industrial Engineering, University of Trento, I-38122 Trento, Italy} \email[M. Cassetta]{michele.cassetta@univr.it} \author{\firstname{Alessandro} \lastname{Pisello}\CDRorcid{0000-0002-7772-4678}} \addressSameAs{5}{Department of Physics and Geology University of Perugia, piazza Universit\`{a}, 06100 Perugia, Italy} \email[A. Pisello]{alessandropisello@gmail.com} \author{\firstname{Diego} \lastname{Perugini}\CDRorcid{0000-0002-2888-6128}} \addressSameAs{5}{Department of Physics and Geology University of Perugia, piazza Universit\`{a}, 06100 Perugia, Italy} \email[D. Perugini]{diego.perugini@unipg.it} \begin{abstract} Experiments were performed at 500 MPa, 1080~\textdegree C and water activities (\textit{a}H$_{2}$O) from 0.0 to 1.0, in fluid-present and fluid-absent conditions, with the aim of constraining the effect of volatiles on phase equilibrium assemblages of a shoshonite from Vulcanello (Aeolian Islands, Italy). Experiments were run both under reducing and oxidizing conditions and results show that proportions, shapes and size of crystals vary as a function of the volatile composition (\textit{X}H$_{2}$O and \textit{X}CO$_{2}$) and volatile content. Clinopyroxene (Cpx) is the main crystallising phase and is compositionally analogous to Cpx crystals found in the natural rock. Plagioclase (Pl) is stable only for water activity lower than 0.1, whereas Fe--Ti oxides are present in all experimental runs, except for those where log \textit{f}O$_{2}$ was lower than ${-}$9, ({${\Delta}$}NNO ${-}$0.11) irrespective of the presence of CO$_{2}$. The addition of CO$_{2}$ (2.8 wt\%) in nominally dry experimental charges substantially reduces the crystallinity by ca. 1/3 compared to volatile free experiments. This result has important consequences upon the physical properties of the magma because it influences its viscosity and, as a consequence, velocity during its travel to the Earth surface. Assuming that the widths of Vulcanello conduits vary from 0.5 to 1.5 m, estimates of the ascent velocity vary in the range 1.5 ${\times}$ 10$^{-4}$--3.5 ${\times}$ 10$^{-2}$ m${\cdot}$s$^{-1}$ for CO$_{2}$ free systems and from 5.7 ${\times}$ 10$^{-4}$--1.3 ${\times}$ 10$^{-1}$ m${\cdot}$s$^{-1}$ for CO$_{2}$ bearing systems. Since shoshonitic magmas are common not only in the Italian volcanic provinces (Aeolian Arc, Campi Flegrei, Ischia Island, Pontine Islands, Monti Cimini, Monte Amiata, Capraia Island, Radicofani, Roccamonfina) but also in different volcanoes worldwide (Yellowstone, Mariana Arc, Kurile Arc, Tonga Arc, Andean Arc, Kamchatka Arc), we suggest that the new data will be useful to better understand shoshonitic magma behaviour under relevant geological scenarios. As such, we also suggest that hazard evaluation should incorporate the probability of very rapid ascent of {poorly-}evolved melts from depth. \end{abstract} \keywords{\kwd{Vulcanello} \kwd{Shoshonite} \kwd{Phase equilibria} \kwd{Mineral assemblage} \kwd{H$_{2}$O--CO$_{2}$}} \maketitle \twocolumngrid \end{noXML} \section{Introduction}\label{sec1} {V}olatile{s} dissolved in ma{gmas are k}ey parameters {to understand} geophysical and geochemical signals {at depth}, as the{ir} exsolution provides {the} {needed} driving force {for magma rise and, possibly, trigger} explosive eruptions. {Being a multiphase system, most magmas are composed of} phenocrysts {c}oexisting with a {volatile bearing} silicate melt (e.g. H$_{2}$O, CO$_{2}$, S species, F, Cl, etc.\,\ldots){. Phase equilibrium studies are of paramount importance to better decipher} magma storage conditions {and physical properties during ascent to the surface. Thus, the role of p}arameters govern{ing} equilibrium between melt and phenocrysts {such as} pressure, temperature, redox state, volatile conten{t, must be known}. {The best way to constrain magma storage conditions of volcanic rocks and changes occurring during ascent} is {to compare the natural solid assemblages} {with} the phase assemblage and compositions obtained {through} phase-equilibrium experiments in which all these parameters are controlled \citep[e.g.,][]{Marteletal2019}. Dissolved volatiles are known to affect the rheological properties of magmas during ascent to the Earth's surface and {H}$_{2}${O is the most important volatile in igneous systems since it affects density, viscosity, as well as phase equilibria, including crystallinity and liquidus temperature \citep[e.g.,][]{Burnham1981,Schulzeetal1997}. It is well known that the amount of dissolved water in melts, as well as the mode of degassing in magma conduits, both affect eruptive styles [e.g., effusive vs. explosive, \citealp{SissonLayne1992}; plinian vs. pelean, \citealp{Marteletal1998}].} However, water is not the only volatile component in magmas. Over tha last decades researchers have focused their efforts on trying to understand how different volatile ratios, such as H$_{2}$O/CO$_{2}$, can influence the magma properties in terms of phase equilibria, mobility of chemical elements, and the amount, shape and size of crystals \citep[e.g.][and references therein]{Kent2008,Marteletal2019}. The presence of CO$_{2}$ reduces water activity and it is commonly assumed that the decreasing H$_{2}$O/CO$_{2}$ ratio in the melt results in an increase of the liquidus temperature and crystal/melt ratio in most silicate systems \citep[e.g.,][]{ClemensWall1981,ScailletEvans1999}. However, the results of some experimental studies seem to indicate that the opposite may occur in basaltic or depolymerized systems at very high CO$_{2}$ concentrations (low water activities) \citep{Giuffridaetal2017}. These results are unexpected and contrast with our common assumption of the role of CO$_{2}$--H$_{2}$O-bearing fluids on phase relationships. \looseness=-1 In order to expand the experimental dataset on the effect of CO$_{2}$ on phase equilibria, we present new H$_{2}$O- and CO$_{2}$-bearing high temperature and high pressure crystallization experiments performed on a shoshonitic composition from the Vulcanello peninsula (Island of Vulcano, Aeolian Island, Italy). Experiments were targeted to investigate the phase stabilities at pressure and temperature conditions relevant for/to some of the southern Italy volcanoes such as Vulcanello and, possibly, to other chemically similar volcanic systems. We discuss experimental results in terms of phase equilibria but we also address how both H$_{2}$O and CO$_{2}$ volatiles affect rheological properties of the magma depending on their content. \subsection{Vulcanello magmatic source and products} {Based on the analysis of f}luid inclusions in quartz xenoliths, \citet{Zanonetal2003} provided evidence that the deepest level of magma storage at Vulcanello is located between 17 and 21 km, close to the Moho [21--25 km; \citealp{Falsaperlaetal1985}] and exclude{d} significant magma ponding at mid-crustal levels. {This is in accordance with} the aeromagnetic data at Vulcano {pointing to a} magma reservoir between 18 and 21 km depth \citep{DeRitisetal2013}. {Moreover,} \citet{Zanonetal2003} {provided and estimate for} Vulcanello magmas {temperature of} 1083 ${\pm}$ 40~\textdegree C. {The formation of Vulcanello occurred during the last Eruptive Epoch of the island of Vulcano [Eruptive Epoch 8, \citealp{DeAstisetal2013}], between AD 1100 to 1250 \citep{Arrighietal2006}, simultaneously to effusive and explosive activity in the La Fossa crater \citep{Fusilloetal2015}.} {The lava platform comprises several superimposed shoshonitic lava flows, which, together with the products of explosive activity that form a scoria cone, were the first subaerial products of the peninsula (Vulcanello 1).} The rocks composing the lava platform are mainly shoshonitic {and} contain large (up to 1.5 cm) xenocrysts of plagioclase \citep{Davietal2009,DeAstisetal2013,Fusilloetal2015}. {After a short time break, the activity resumed with explosive eruptions (Vulcanello 2) which led to the formation of a second cone during a single eruptive unit, showing 3-m-thick fallout deposit having shoshonitic composition \citep{Fusilloetal2015}. Following Vulcanello 2, no activity took place during ${\sim}$500~years. Subsequently, lava emission resulted in the emplacement of the latitic Roveto lava flow, that was followed by three explosive phases leading to the edification of Vulcanello 3 cone. These} rock{s are} latitic {in} composition {with the presence of} clinopyroxene{s} megacrysts [ca. 1.5 cm; \citealp{Fusilloetal2015}]. {The activity of Vulcanello 3 ended with the emplacement of a final lava flow (Valle dei Mostri} latitic lava {flow), which flowed in the same direction than the Roveto} latitic lava flow \citep{Fusilloetal2015,Nicotraetal2018,Davietal2009}. \vspace*{-5pt} \section{Experimental and analytical methods}\label{sec2} \subsection{Starting material and experiments}\label{sec2.1} The starting material of our experiments is a natural scoria lava from Vulcanello 1. The bulk rock\break composition (Table~\ref{tab1}) is characterized by SiO$_{2} = 53.40$ wt\% and Na$_{2}$O $+$ K$_{2}$O $=$ 8.76 wt\%. In the total Alkali-Silica (TAS) diagram this composition plots in the field of shoshonite ({Table~\ref{tab1}}). The modal abundance of phenocrysts, mainly composed of clinopyroxene (Cpx) and plagioclase (Pl), is 8--11~vol\%. The phenocrysts often contain inclusions of Fe--Ti oxides. As reported in \citet{Davietal2009} the composition of clinopyroxene {and feldspar} phenocrysts {ranges between} Wo$_{49\text{--}44}$En$_{36\text{--}42\text{--}36}$Fs$_{9\text{--}20}$ {and} An$_{5\text{--}46}$Ab$_{51\text{--}52}$Or$_{3\text{--}43}$, respectively. Feldspar microphenocrysts show a {similar} composition to that of phenocrysts (An$_{3\text{--}35}$Ab$_{51\text{--}61}$Or$_{4\text{--}46}$). {O}livine {has a compositional range} Fo$_{42\text{--}62}$ {and t}he Fe--Ti oxides {have} Usp content {close to} 14 mol\%. The groundmass consists of glass, leucite (Lc), Fe--Ti oxides, Cpx, Pl and small amount of olivine (Ol). \begin{table} \caption{\label{tab1}Compositions of starting glass} \tabcolsep=7pt \begin{tabular}{cccc} \thead & \parbox[t]{1cm}{\centering VL0 (wt\%)} & std & \parbox[t]{2.5cm}{\centering \citet{Vetereetal2007} (wt\%)}\vspace*{2pt} \\ \endthead $n$ & 100 & & 100 \\ SiO$_{2}$ & 53.40 & 0.41 & 53.47 \\ TiO$_{2}$ & 0.70 & 0.04 & 0.71 \\ Al$_{2}$O$_{3}$ & 15.78 & 0.24 & 15.48 \\ FeO & 8.04 & 0.36 & 8.39 \\ MnO & 0.10 & 0.12 & 0.10 \\ MgO & 4.88 & 0.15 & 4.88 \\ CaO & 8.81 & 0.22 & 8.51 \\ Na$_{2}$O & 3.66 & 0.23 & 3.66 \\ K$_{2}$O & 3.71 & 0.13 & 4.72 \\ P$_{2}$O$_{5}$ & 0.62 & 0.21 & \\ LOI\vspace*{5pt} & & & \\ Total & 99.76 & & 99.92 \botline \end{tabular} \tabnote{Notes: Oxide components are given in wt\%. Errors represent one standard deviation (std). Shoshonite studied by \citet{Vetereetal2007} is also reported for comparison.} \end{table} Sample preparation prior to high-pressure experiments was performed at the Petro-Volcanology Research Group laboratories of the University of Perugia (PVRG; \url{http://pvrg.unipg.it}). About 100 g of crushed bulk rock was melted in a Pt crucible in air at 1600~\textdegree C for 4 h in a Pt$_{80}$Rh$_{20}$ crucible. Melting was performed in a Nabertherm HT 04/17 MoSi$_{2}$-heated box furnace. The melt was quenched to a glass by pouring it on a brass plate. The glass was crushed, re-melted and quenched again using the same procedure. This technique ensured compositional homogeneity of the glass and avoided crystallization \citep{Vetereetal2015a}. The composition of the glass and the absence of crystals were controlled by electron microprobe imaging and measurements with a CAMECA SX 100 at the Institute of Mineralogy University of Hannover (see details below;\break {Table~\ref{tab1}}). {The} NBO/$T$ parameter corresponds to the number of non-bridging oxygens (NBO) per tetrahedrally coordinated cation ($T$) and {is a good proxy to depict the structure of silicate melts in terms of polymerization} \citep{MysenRichet2005}. This ratio can be calculated as follows: {\begin{equation} \label{eq1} \frac{\mathrm{NBO}}{T}=\frac{1}{T}\times {\sum} _{i=1}^{i}n\times {M}_{i}^{n+} \end{equation}}\unskip where $T$ is the total atomic abundance of tetrahedrally coordinated cations, $M$ is the proportion of network modifying cations $i$, with electrical charge $n+$ after subtraction of the portion required for charge-balancing trivalent cations on tetravalent sites \citep{MysenRichet2005}. Our starting composition has a NBO/$T$ of 0.39. The fragility, $F$, of our starting composition, was also calculated and has a value of 0.60. This parameter is useful to estimate the deviation from Arrhenian behaviour upon cooling in a plot relating viscosity and the scaled temperature ($T_{g}/T$) while approaching the glass transition temperature $T_{g}$ \citep{Angell1995,GiordanoDingwell2003}. Both the NBO/$T$ and $F$ point to a high degree of depolymerisation of our starting material. In order to prepare volatile-bearing charges, we used the following procedure: The capsules (Au$_{80}$Pd$_{20}$; 5 mm diameter, 3 cm in length, previously annealed at 1{100~\textdegree C} for 10 min and then welded on one end) w{ere} filled {with the} following procedure: ({a}) the capsule {was} charg{ed} with the desired amounts of {${\pm}$}H$_{2}$O {${\pm}$} Ag$_{2}$C$_{2}$O$_{4}$ {+} glass powder (Ag$_{2}$C$_{2}$O$_{4}$ {is the} source for CO$_{2}$ {as the} silver oxalate decomposes during heating and generates {the} carbon dioxide); ({b}) a steel piston {was used to gently} compact the charge; ({c}) {to avoid any volatiles release,} the capsule was tightly squeezed on top, rolled up in a wet tissue, and frozen by placing it into a bath of liquid nitrogen; ({d}) the upper end of the capsule was welded shut with a conventional graphite-arc welder. {In agreement with literature data on the Vulcanello magmatic source (21--17 km depth and} 1083 ${\pm}$ 40~\textdegree C, {see above) we performed experiments at 500 MPa and 1080~\textdegree C for 70 h, in order to shed light on processes relevant to the magmatic source of Vulcanello magma.} \begin{figure}[t!] %\vspace*{-2pt} \includegraphics{fig01} %\vspace*{-2pt} \caption{\label{fig1} H$_{2}$O and CO$_{2}$ concentrations added to the capsules before experiment (at 500 MPa and 1080~\textdegree C). The shaded area represent the undersaturated conditions in accord with the modeled H$_{2}$O and CO$_{2}$ concentrations at fluid saturation obtained by \citet{Vetereetal2014,Vetereetal2011} for the same composition is also shown (solid black curve). Orange and blue symbols indicate experiments performed at reduced and oxidized conditions, respectively, (see Table~\ref{tab3}). Green and black circles are nominally anhydrous VL46 and 47 experiments, respectively (Table~\ref{tab3}). Red dot refers to the sample with the highest water content.} %\vspace*{-2pt} \end{figure} \begin{table*} \caption{\label{tab2}Results of microprobe analyses of residual glasses} \tabcolsep=8.3pt\fontsize{9.6}{11.52}\selectfont \begin{tabular}{lcccccccc} \thead & VL32 & std & VL33 & std & VL34 & std & VL36 & std \\ \endthead SiO$_{2}$ & 54.86 & 0.39 & 55.53 & 0.47 & 54.28 & 0.60 & 52.17 & 0.55 \\ TiO$_{2}$ & \00.71 & 0.05 & \00.73 & 0.04 & \00.73 & 0.06 & \00.71 & 0.04 \\ Al$_{2}$O$_{3}$ & 20.61 & 0.26 & 20.71 & 0.34 & 19.87 & 0.22 & 17.99 & 0.23 \\ FeOT & \01.81 & 0.57 & \01.21 & 0.27 & \01.99 & 0.31 & \02.95 & 0.18 \\ MnO & \00.10 & 0.06 & \00.09 & 0.04 & \00.15 & 0.09 & \00.13 & 0.11 \\ MgO & \01.72 & 0.15 & \01.51 & 0.57 & \02.13 & 0.37 & \03.07 & 0.46 \\ CaO & \04.70 & 0.31 & \04.13 & 0.27 & \05.16 & 0.27 & \06.23 & 0.10 \\ Na$_{2}$O & \06.32 & 0.26 & \06.66 & 0.16 & \06.08 & 0.21 & \05.42 & 0.17 \\ K$_{2}$O & \04.91 & 0.39 & \05.11 & 0.21 & \04.82 & 0.28 & \04.25 & 0.32 \\ P$_{2}$O$_{5}$ & \00.49 & 0.14 & \00.56 & 0.21 & \00.45 & 0.10 & \00.46 & 0.16\vspace*{5pt} \\ Total & 96.23 & & 96.24 & & 95.66 & & 93.38 & \\ H$_{2}$O & \03.77 & & \03.76 & & \04.34 & & \06.62 & \\ \hline & VL37 & std & VL38 & std & VL39 & std & VL41 & std \\ \hline SiO$_{2}$ & 55.50 & 0.68 & 56.03 & 0.47 & 55.72 & 0.91 & 53.90 & 0.21 \\ TiO$_{2}$ & \00.45 & 0.13 & \00.49 & 0.10 & \00.49 & 0.09 & \00.62 & 0.05 \\ Al$_{2}$O$_{3}$ & 20.35 & 1.03 & 20.67 & 0.62 & 20.18 & 0.72 & 18.57 & 0.27 \\ FeOT & \04.32 & 1.07 & \04.12 & 0.70 & \03.96 & 0.58 & \04.64 & 0.23 \\ MnO & \00.12 & 0.05 & \00.09 & 0.10 & \00.09 & 0.10 & \00.10 & 0.11 \\ MgO & \01.29 & 0.38 & \01.14 & 0.23 & \01.10 & 0.32 & \02.00 & 0.37 \\ CaO & \03.99 & 0.40 & \03.69 & 0.42 & \03.60 & 0.32 & \04.66 & 0.17 \\ Na$_{2}$O & \06.65 & 0.35 & \06.74 & 0.38 & \06.95 & 0.17 & \05.87 & 0.10 \\ K$_{2}$O & \04.86 & 0.27 & \05.04 & 0.24 & \05.07 & 0.26 & \04.65 & 0.18 \\ P$_{2}$O$_{5}$ & \00.57 & 0.23 & \00.51 & 0.16 & \00.50 & 0.17 & \00.46 & 0.12\vspace*{5pt} \\ Total & 98.10 & & 98.52 & & 97.66 & & 95.47 & \\ H$_{2}$O & \01.90 & & \01.48 & & \02.34 & & \04.53 & \\ \hline & VL45 & std & VL46 & std & VL47 & std & & \\ \hline SiO$_{2}$ & 51.36 & 0.99 & 56.02 & 0.83 & 55.69 & 0.44 & & \\ TiO$_{2}$ & \00.63 & 0.02 & \00.76 & 0.02 & \00.75 & 0.04 & & \\ Al$_{2}$O$_{3}$ & 15.01 & 0.24 & 18.10 & 0.34 & 18.55 & 0.64 & & \\ FeOT & \07.29 & 0.41 & \07.15 & 0.46 & \06.75 & 0.41 & & \\ MnO & \00.10 & 0.10 & \00.13 & 0.08 & \00.18 & 0.09 & & \\ MgO & \04.19 & 0.29 & \02.40 & 0.10 & \02.23 & 0.39 & & \\ CaO & \07.34 & 0.44 & \04.33 & 0.12 & \04.48 & 0.66 & & \\ Na$_{2}$O & \02.81 & 0.39 & \06.07 & 0.26 & \06.21 & 0.20 & & \\ K$_{2}$O & \03.61 & 0.13 & \04.01 & 0.08 & \04.01 & 0.27 & & \\ P$_{2}$O$_{5}$ & \00.51 & 0.04 & \00.74 & 0.03 & \00.75 & 0.07\vspace*{5pt} & & \\ Total & 91.85 & & 99.70 & & 99.61 & & & \\ H$_{2}$O & \08.15 & & \00.30 & & \00.39 & & & \botline \end{tabular} \xxtabnote{Notes: H$_2$O represent the water content estimated by difference methods. Microprobe data are reported for all the residual glasses investigated in wt\% together with errors represented by one standard deviation (std).} \end{table*} %\onecolumngrid \begin{sidewaystable} \vsptab \caption{\label{tab3}Phase assemblage results derived from image analyses methods} \tabcolsep=4.5pt \begin{tabular}{cccccccccccccccc} \thead & \morerows{1}{\parbox[t]{.8cm}{\centering H$_{2}$O$_{\mathrm{in}}$ (wt\%)}} & \morerows{1}{\parbox[t]{.8cm}{\centering CO$_{2}$ (wt\%)}} & \xmorerows{1}{\textit{a}H$_{2}$O\tsup{\pounds}} & \xmorerows{1}{log \textit{f}O$_{2}$} & \xmorerows{1}{${\Delta}$QFM} & \multicolumn{4}{c}{CPx} & \multicolumn{4}{c}{Plg} & Ox & \morerows{1}{\parbox[t]{1cm}{\centering $\Phi_{\mathrm{tot}}$ (area\%)}} \\ \cline{7-10} \cline{11-14} & & & & & & Fs & Wo & En & (area\%) & An & Or & Ab & $\Phi_{\mathrm{Plg}}$ (area\%) & $\Phi_{\mathrm{Ox}}$ (area\%) & \\ \endthead VL32\tsup{\#} & \03.5 & 0.651 & 0.36 & ${-}$10.80 & ${-}$1.37 & 17.31 & 46.40 & 36.28 & 16.1 & & & & & & 16 \\ VL33\tsup{\#} & \03.5 & 0.291 & 0.35 & ${-}$10.82 & ${-}$1.39 & 13.25 & 47.03 & 39.71 & 16.3 & & & & & & 16 \\ VL34\tsup{\#} & \04.2 & 0.092 & 0.44 & ${-}$10.64 & ${-}$1.21 & 16.11 & 45.87 & 38.00 & 14.3 & & & & & & 14 \\ VL35\tsup{\#} & \05.5 & 0.919 & 0.61 & ${-}$10.35 & ${-}$0.92 & & & & 10.7 & & & & & & 11 \\ VL36\tsup{\#} & \06.5 & 0.558 & 0.74 & ${-}$10.18 & ${-}$0.75 & & & & 10.2 & & & & & & 10 \\ VL37 & \02.0 & 0.693 & 0.14 & \0${-}$7.64 & \mn1.79 & & & & 15.8 & & & & & 1.7 & 17 \\ VL38 & \01.5 & 0.297 & 0.09 & \0${-}$8.01 & \mn1.42 & & & & 18.1 & & & & & 1.7 & 22 \\ VL39 & \02.5 & 0.070 & 0.19 & \0${-}$7.38 & \mn2.05 & & & & 18.7 & & & & & 1.9 & 21 \\ VL40 & \03.0 & 0.805 & 0.31 & \0${-}$6.94 & \mn2.49 & 15.67 & 41.34 & 42.96 & 17.0 & & & & & 1.5 & 18 \\ VL41 & \04.5 & 0.620 & 0.48 & \0${-}$6.57 & \mn2.86 & 16.45 & 47.85 & 35.68 & 14.0 & & & & & 1.4 & 15 \\ VL45\tsup{\$} & 13.0 & 0.000 & 1.00 & \0${-}$4.94 & \mn4.49 & & & & & & & & & & \03 \\ VL46* & \00.0 & 2.830 & 0.01 & ${-}$10.26 & ${-}$0.83 & 17.33 & 42.91 & 39.75 & 15.3 & 33.58 & 9.60 & 56.82 & 5.9 & & 21 \\ VL47* & \00.0 & 0 & 0.02 & \0${-}$8.92 & \mn0.51 & 17.90 & 42.42 & 39.68 & 21.6 & 32.91 & 9.40 & 57.69 & 8.8 & & 30 \botline \end{tabular} \xxxtabnote{Crystals composition are also reported.} \xxxtabnote{Note: Crystallinity ($\Phi$) is reported in vol\% and derived from 500 MPa and temperature of 1080~\textdegree C experiments. H$_2$O and CO$_2$ are reported in wt\%. Water activity is also reported for all run while phase compositions are reported for most of the experimental runs except for those were crystal size was too small in order to be analyses. \tsup{\#}Denote experiments performed at controlled \textit{f}O$_2$. *Denote experiments were no water was added to the charges. \tsup{\$}Denote experiments with quench crystals. \tsup{\pounds}Water activity is calculated from the amount of water estimated by the difference method (see Table~\ref{tab2}).} \end{sidewaystable} %\twocolumngrid A total of 13 experiments with variable volatile contents (see {Figure~\ref{fig1}}) were performed at 500 MPa and 1080~\textdegree C ({Tables~\ref{tab2}} and {\ref{tab3}}). Experiments were done in an internally heated pressure vessel (IHPV) equipped with a rapid-quench sample holder at the Institute fur Mineralogy, Leibniz University of Hannover (Germany). Details for the IHPV are provided in \citet{Berndtetal2002}. Temperature was measured using four unsheathed S-type thermocouples (precision of temperature was ${\pm}$5~\textdegree C). The IHPV was pressurized with argon. Pressure \pagebreak was monitored\unskip\break \phantom{~}\break\clearpage\noindent using digital pressure transducers having an uncertainty of about 1 MPa. The variation of pressure during the experiments was ${\leq}$5 MPa. Some experiments were performed at the intrinsic oxygen fugacity (\textit{f}O$_{2}$) conditions of the pressure vessel which was found to be close to QFM+4 (the intrinsic \textit{f}O$_{2}$ corresponding to H$_{2}$O-saturated conditions), with QFM being the quartz--fayalite--magnetite equilibrium \citep{Berndtetal2002}. At H$_{2}$O-undersaturated conditions, \textit{f}O$_{2}$ is lower, depending on the prevailing water activity [QFM${-}$0.8 to QFM$+$3; \citealp{Botcharnikovetal2005}]. In our experiments \textit{f}O$_{2}$ was calculated using the following relation: {\begin{equation}\label{eq2} \log f\mathrm{O}_{2}=\log f\mathrm{O}_{2} (\text{at } X\mathrm{H}_{2}\mathrm{O}_{\mathrm{in}}=1) - 2 \log X\mathrm{H}_{2}\mathrm{O}_{\mathrm{in}}. \end{equation}}\unskip \citep{ScailletEvans1999}. We estimate that the maximum error on the calculated \textit{f}O$_{2}$ is about 0.2--0.3 log units according to \citet{Botcharnikovetal2005}. Other experiments were performed under controlled \textit{f}O$_{2}$ conditions (QFM${-}$0.7 to QFM${-}$1.4) by adding H$_{2}$ to the Ar pressure medium. The hydrogen fugacity (\textit{f}H$_{2}$) in these experiments was monitored using a Shaw membrane as described in \citet{Berndtetal2002}. The \textit{f}H$_{2}$ was adjusted to be \textit{f}H$_{2}=20$ bar at experimental P--T conditions. Experiments were drop quenched with a cooling rate of approximately 150--200~\textdegree C/s \citep{Berndtetal2002}. For one experiment ({Table~\ref{tab3}}), it was not possible to avoid the formation of quench crystals due to the relatively high amount of added water (see results for details). {\vspace*{-2pt}} \subsection{Electron microprobe and image analyses}\label{sec2.2} All run-products were mounted in epoxy, ground flat and polished for textural and chemical analyses. Glasses and minerals produced during the experiments were analysed using a Cameca SX100 microprobe at the Institute of Mineralogy. An accelerating voltage of 15~kV was used with a beam current of 15~nA for silicate minerals and oxides and 6~nA for glasses. The beam was defocused to at least 15~${\upmu}$m for glass analyses and focused to ${<}$2 ${\upmu}$m for the analysis of crystal phases. All elements were analysed with a counting time of 10 s on peak. Standards used for calibration were Fe$_{2}$O$_{3}$, MgO, MnTiO$_{3}$ (Mn and Ti), albite (Na), wollastonite (Si for mineral and glass and Ca for glass), apatite (P and F for glass and F, Ca and P for apatite), orthoclase (K), anhydrite (Ca and S). Standard deviations reported in Table~\ref{tab2} were\break calculated based on 10 to 20 analysed spots on the same sample. Raw data were corrected with the software ``Peak Sight'' and ``PAP'' matrix \citep{PouchouPichoir1991}. Precision and accuracy were determined by measuring reference glasses VG-568 (rhyolite) and VG-2 (basalt) from Smithsonian standards collection \citep{Jarosewich2002}. Due to the presence of crystals in experimental charges, the water content of the glasses was estimated following the ``by-difference'' method described in \citet{Devineetal1995}. Following \citet{Paratetal2008}, we estimate that the error on determination of H$_{2}$O content is ${\sim}$0.5 wt\%. CO$_{2}$ in glasses were not determined but estimated from the solubility curve in Figure~\ref{fig1}. Back-scattered electron (BSE) images were collected with a Philips FEG (Field Emission Gun) Quanta 200F equipped with a Si/Li-SUTW detector (EDAX, Philips Electronics) installed at the University of Calabria (Italy). Representative BSE images were collected at different magnifications (400$\times$ up to 1600$\times$) depending on the crystal size. Surface percentage of the different phases was determined for all images using grey levels with Image-ProPlus 6.0 \citep[see details in][]{Vetereetal2015a}. Seven to ten BSE images obtained in different parts of the same section were evaluated for homogeneity. We quantified surface fractions of glass, clinopyroxene, spinel, plagioclase and gas bubbles. In total, 110 BSE images were analysed. \section{Results}\label{sec3} The crystalline phases observed in our experiments are clinopyroxene (Cpx), plagioclase (Pl), and Fe--Ti oxides (Ox) in various proportions, depending on the fraction of volatiles dissolved in the melt. The results show that Cpx is the main phase with abundance increasing from 10.2\% to 21.6\% as \textit{a}H$_{2}$O decreases from 0.74 to nearly zero (Table~\ref{tab3}). Water activity also plays a major role on the stability of Pl that is only present in experiments with a melt H$_{2}$O content ${\leq}$1 wt\% (Table~\ref{tab3}). Plagioclase has a maximum abundance of 8.8\% in the volatile-free sample (VL 47$^{*}$ in Table~\ref{tab3}). Minor Fe--Ti oxides (${\leq}$2\%) are present in samples equilibrated under relatively high oxygen fugacity conditions. Based on the amounts of added volatiles and in agreement with solubility model and data\break \citep[e.g.][]{Vetereetal2014} one can distinguish between fluid-saturated (fluid present) and fluid-undersaturated (fluid absent) runs with respect to H$_{2}$O $+$ CO$_{2}$ ({Figure~\ref{fig1}}). Quenched melt (residual glass) is present in all run products with a mode ranging from 70\% in VL47$^{*}$ to 97\% in VL45\tsup{\$} (Table~\ref{tab3}). This wide range of melt fraction at constant pressure and temperature shows the importance of water activity (\textit{a}H$_{2}$O) on the liquidus temperature and crystallinity. All glass compositions, including the starting composition, are listed in Tables~\ref{tab1} and~\ref{tab2} and are illustrated in {Figure~\ref{fig2}}. The compositions of residual glasses range from 51 to 56 wt\% SiO$_{2}$ (Table~\ref{tab2} and Figure~\ref{fig2}). Glass compositions evolve as a function of the degree of crystallinity. We however note that a larger crystallinity does not always correspond to an enrichment of SiO$_{2}$ in the residual glass (compare Table~\ref{tab2} and 3). We also note that in the experiment performed under oxidizing and water-saturated (no CO$_{2}$) conditions, the conditions are close to the liquidus and the melt has a composition very similar to that of the starting glass (although comparing alkali content in the VL45 and starting material we do note a difference of ca. 0.5 wt\% of Na$_{2}$O possibly due to the EPMA measurements conditions; red point in Figures~\ref{fig1} and~\ref{fig2}). Experiments performed at reducing conditions and containing both water and CO$_{2}$ show a SiO$_{2}$ trend that increases nearly linearly as the MgO content decreases. The same observation can be done for, Na$_{2}$O and K$_{2}$O. The total FeO decreases with decreasing MgO but FeO concentrations are always higher in oxidized experiments than in reduced experiments at a given MgO content, suggesting a possible Fe-loss to the capsule wall at reducing conditions, as discussed below. TiO$_{2}$ is relatively constant in reduced experiments but decreases with decreasing MgO in oxidized runs. Due to the crystallization of the main phase Cpx, CaO continuously decreases with MgO (with increasing crystallinity). \begin{figure*} \includegraphics{fig02} \caption{\label{fig2} Change of the residual glass composition (in wt\% of element concentrations) in the silicate melts. The analyses are recalculated to a total of 100\%. The arrow on the CaO vs. MgO plot indicates the vector pointing to the Cpx composition. The grey triangle refers to the starting material composition while reddish and bluish circles indicate experiments performed at oxidized and reduced conditions, respectively, as reported in Tables~\ref{tab2} and~\ref{tab3}. Red dot as in Figure~\ref{fig1}.} \end{figure*} Eight experiments were performed at volatile saturated conditions and six at volatile undersaturated conditions (Figure~\ref{fig1}). {Figure~\ref{fig3}} shows the variation of crystal content as a function of water content. The general trend indicates that the higher the amount of water in the melt, the lower the amount of crystals, as expected from the numerous previous phase relationships conducted so far. No significant difference between fluid-absent and fluid-present runs is detected in terms of crystal content in samples with an amount of water added to the system that is larger than 1.0 wt\%. Thus, we did not detect a substantial difference in crystal content between experiments performed at fluid-saturated and fluid-undersaturated conditions if the \textit{a}H$_{2}$O is similar. For example, we investigated the effect of increased CO$_{2}$ concentration in water-rich samples (i.e. up to 3.7~wt\% H$_{2}$O; sample VL32 and VL33, see Table~\ref{tab3}) by performing two experiments with a total amount of CO$_{2}$ from 6500 ppm (fluid-saturated) and 2900 ppm (fluid-undersaturated). No difference between CO$_{2}$-rich and CO$_{2}$-poor samples was observed in terms of crystallinity. In particular, both melts showed Cpx contents of ${\sim}$16\% (Table~\ref{tab3}). In contrast, our experiments show that, {at} {very low H}$_{2}${O content}, the presence of CO$_{2}$ causes a substantial decrease of crystal content. A difference up to 10 {area}\% in crystal content is observed between samples VL46 and VL47 (see Table~\ref{tab3}), where the first charge contains a large amount of CO$_{2}$ (2.8 wt\%, nominally dry, no water added) while VL47 is nearly volatile free. It is emphasized that th{is} difference of {10 area}\% in crystallinity determined by image analysis is also consistent with mass balance calculations. Using {the GeoBalance program [}Excel VBA program for mass balance calculation; \citealp{Lietal2020}] the residual melts in VL47 can be best reproduced if 6 {wt}\% {Pl} and 30 {wt}\% {Cpx} crystallize from the starting material (the mineral compositions from experimental products were considered for the calculation, see Table~\ref{tab3}). For VL46, mass balance calculation indicates the crystallization of 3 {wt}\% {Pl} and 25 {wt}\% {Cpx}, point{ing} to the possible role of CO$_{2}$ on phase proportions as discussed\break below. \begin{figure} \includegraphics{fig03} \caption{\label{fig3} Evolution of crystallinity vs. water content of experiments presented in Table~\ref{tab3}. Note the variation on crystallinity between CO$_{2}$ bearing and CO$_{2}$ free sample in nominally dry experimental charges. Symbols as in Figure~\ref{fig1}.} \end{figure} {The a}lphaMELTS code was used to {further} {elucidate} the phase evolution {observed} in {this study.} The alphaMELTS (V. 1.9) software provides a simple text-based interface to subroutine versions of the algorithms MELTS \citep{AsimowGhiorso1998,SmithAsimow2005,Ghiorsoetal2002,GhiorsoSack1995}. {It} allows to calculate equilibrium assemblages along a thermodynamic path set by the user (details {are provided by}: \url{http://melts.ofm-research.org/} and \url{https://magmasource. caltech.edu/forum/}). {Results of the simulations are shown in} {{Figure~\ref{fig4}}} {comparing the crystals content measured via Image Analyses and calculated by using aplhaMELTS. There is a good correlation between calculated and \mbox{measured} data, although VL46 sample shows a calculated crystallinity higher than 40 vol\% (expected from equilibrium thermodynamic calculation) whereas the experimentally observed crystallinity is much lower (Figure~\ref{fig4} and Table~\ref{tab3}).} \begin{figure} \includegraphics{fig04} \caption{\label{fig4} Comparison between Image Analyses measurements and calculat amount of crystals by alphaMELTS approach. Note the good correlation for most of the sample. VL46 sample, instead shows a large departure relatively to the 1:1 line (green circle).} \end{figure} \section{Discussion}\label{sec4} \subsection{Effect of volatiles on phase equilibria}\label{sec4.1} In order to test for equilibrium condition in our experiments we can refer to models proposed by {\citet{Putirka2008}}. {Tests for equilibrium between Cpx and a coexisting liquid can be made by comparing observed and predicted values for Fe--Mg exchange, or K}$_{D}${(Fe--Mg)}$^{\text{cpx--liq}}${, which should be 0.28 $\pm$ 0.08 for Cpx (Putirka,2008; considering FeO$=$FeO}$_{\mathrm{tot}}${). Our results show} values ranging from 0.236 and 0.248 when using compositions in Table~\ref{tab2}, indicating compositions close to equilibrium. The glasses of our isothermal experiments show geochemical trends that are entirely controlled by the degree of crystallinity that itself depends on the composition and amount of volatiles in the system. The evolution trends for most of the elements (Ca, Mg, Si, Al) are consistent with the crystallization of Cpx (Figure~\ref{fig2}), which predominates the solid phase. Ti is almost constant in experiments performed at reduced conditions whereas its abundance is lower in the experiments performed under oxidized conditions, due to the crystallization of Fe--Ti oxides in the latter. The composition of melts from two experiments (VL46 and VL47 at very low H$_{2}$O content; Tables~\ref{tab2} and~\ref{tab3}) deviates slightly from the general trend for some elements and can be explained by the presence of Pl in the run products ({Figure~\ref{fig5}}). \begin{figure*} \vspace*{2pt} \includegraphics{fig05} \caption{\label{fig5} BSE images of sample VL46 and VL47 (see Table~\ref{tab3}) after annealing at 1080~\textdegree C and pressure of 500 MPa. In particular, as 2.8 wt\% CO$_{2}$ is added to the system, the crystal content decreases by about 10 area\%, and also the relative amount of single crystal phases (Cpx and Pl) decreases (VL46*, Table~\ref{tab3}). Clinopyroxene amount varies from 21.6 to 15.1 area\% while plagioclase slightly decreases from 8.8 to 5.9 area\% (VL46* and VL47*, Table~\ref{tab3}).} \vspace*{5pt} \end{figure*} The amount of CO$_{2}$ that can be dissolved in melts is lower than that of H$_{2}$O and the solubility of both volatiles has been investigated as a function of pressure in a variety of melt compositions \citep[e.g.,][]{Brookeretal2001,KingHolloway2002,HollowayBlank1994, Dixonetal1995a,Dixonetal1995b,Jakobsson1997,Morizetetal2010, BakerAlletti2012,Schanofskietal2019}. Carbon species (CO$_{2}$, CH$_{4}$, and CO) have low solubility in basalt and shoshonite melts [at 500 MPa ${<}$ 4500 ppm CO$_{2}$, \citealp{Shishkinaetal2014,Vetereetal2014,Behrensetal2009,Botcharnikovetal2006}] and carbonate minerals are not stable in mafic melts. It could therefore be anticipated that (for low CO$_{2}$ melt concentrations) CO$_{2}$ has no effect on phase compositions and proportions when compared to CO$_{2}$-free dry systems. Indeed, samples with high water contents do not show any important differences in terms of crystallinity between CO$_{2}$-bearing and CO$_{2}$-free samples (Figure~\ref{fig3}). This indicates that water plays the most important role on controlling the crystallization temperatures and that the role of minor to moderate CO$_{2}$ concentrations is subordinate. However, for systems with low water and relatively high CO$_{2}$ levels, this observation seems not to be valid. The difference of 10\% in crystallinity found between nominally dry samples (fluid absent and pure CO$_{2}$ fluid) implies that CO$_{2}$ does not behave as a fully inert component, only reducing water activity. Following the determinations of \citet{Husenetal2016}, it is emphasized that the melts synthesized at nominally dry fluid-absent conditions in the internally-heated pressure vessel used at Hannover probably contain at least 0.5 wt\% H$_{2}$O. Absolutely dry conditions can never be realized because entrapment of adsorbed water during capsule preparation cannot be avoided and because hydrogen can diffuse through the capsule material at high temperature. This process is partially related to reduction of ferric iron to ferrous iron in the melt during the high $T$--$P$ experiments following the reaction {\begin{equation}\label{eq3} \mathrm{H}_{2}({g}) + \mathrm{Fe}_{2}\mathrm{O}_{3}({m}) = 2 \mathrm{FeO}({m}) +\mathrm{H}_{2}\mathrm{O}({m}) \end{equation}}\unskip where \textit{m} refers to the melt phase and \textit{g} the gas phase. Thus, the difference between fluid-absent and fluid-present experiments can be considered as a minimum value and may be even larger than determined in this study, if absolutely H$_{2}$O-free systems are\break considered. Although experimental conditions are very different, our observations are in agreement with\break observations of \citet{Fiegeetal2015} showing that the addition of ${\sim}$2000 ppm of CO$_{2}$ to a hydrous basaltic melt (5 wt\% H$_{2}$O) has a small effect on crystal content (Cpx, Spl) and glass fractions (during decompression). Nevertheless, when evaluating in detail those data, we can see that samples containing only H$_{2}$O show relatively higher abundance of Cpx compared to those containing both H$_{2}$O and CO$_{2}$. Experiments simulating the effects of CO$_{2}$ fluxing conducted by \citet{Giuffridaetal2017} also indicate that the addition of CO$_{2}$ to a partially crystallized basalt can influence the phase proportion and result in a partial dissolution of Cpx, in contrast to the common assumption. \citet{Vetereetal2015b} noted that the crystallinity of Etna basalt decreases as CO$_{2}$ is added to the system, which is in agreement with the data from this study on a shoshonitic system. In particular, in those experiments, Pl was the most abundant mineral phase, but Pl was absent when only CO$_{2}$ was added to the experimental charge. In addition, CO$_{2}$ appeared to lower Cpx abundance \citep{Vetereetal2015b}. Thus, CO$_{2}$ may not act as a complete inert phase in mafic systems in which it is mainly incorporated as carbonate\break species. A closer look on data presented in Figure~\ref{fig3} allows us to further discuss the effect of the CO$_{2}$ in magmatic systems. In fact, Figure~\ref{fig3} shows, at least for added water content higher than 2 wt\%, a nearly linear correlation between crystallinity and added water content, independently on \textit{f}O$_{2}$. In {Figure~\ref{fig6}} we show that, if the amount of added CO$_{2}$ is larger than 0.65~wt\% and added water content is close or lower than 2 wt\%, the crystallinity of the run products becomes lower than what would be predicted from the nearly linear trend between crystallinity and added water, as highlighted by the dashed black and red lines (Figure~\ref{fig6}). \begin{figure} %\vspace*{-3pt} \includegraphics{fig06} \caption{\label{fig6} {Insight on the evolution of crystallinity vs. water content of experiments presented in Table~\ref{tab3}. Note the variation on crystallinity (red and black dashed curves) between samples having water content lower than 3.5~wt\% with CO$_{2}$ ${>}$ 0.65 wt\% and CO}$_{2} < 0.65$ wt\%. Equations curves are also reported.} %\vspace*{-3pt} \end{figure} More in details, the flushing experiment performed by \citet{Giuffridaetal2017}, showed that a CO$_{2}$-rich fluid phase leads to an increase of the amount of Cpx and a decrease of the abundance of Pl at 300 MPa. This decrease of Pl proportion is associated with a change in An content. Although the conditions are very different, as described above, this proves an effect of CO$_{2}$ on phase stability. {Even if small, we also observed a change in An content from CO}$_{2}$ {free to CO}$_{2}$ {bearing (VL46 to VL47) as observed in} \citet{Giuffridaetal2017} from high to low pressure experiments on Etna basalt. The glass compositions presented in Table~\ref{tab2} also highlight some differences between VL46 and VL47 samples. The SiO$_{2}$ and Al$_{2}$O$_{3}$ concentrations are lower in VL46 whereas FeO and CaO concentrations are higher in VL46. This leads to a small variation of NBO/$T$ (0.33 to 0.28) but could mirror the small mineral chemistry variations that we observed. Finally, a liquidus temperature depression, caused by dissolved CO$_{2}$ in melt, is expected and this depression may range from negligible to several hundred degrees depending on silicate melts composition and pressure \citep{Eggler1975,Eggler1976,MysenBoettcher1975, EgglerKadik1979, EgglerRosenhauer1978, Boettcheretal1987, Boettcher1984}. The extent of this depression is linked to the solubility and solubility mechanism(s) of CO$_{2}$ in the melt. For instance, the liquidus temperature depression of the CaMgSi$_{2}$O$_{6}$--CO$_{2}$ system compared to the volatile free system illustrates this mechanism \citep{EgglerRosenhauer1978, MysenRichet2005}. Finally, it is emphasized that the difference in liquidus temperature for the CO$_{2}$ bearing and CO$_{2}$ free compositions (samples VL46 and VL47) observed experimentally at 500 MPa could not be reproduced with alphaMELTS, with $T_{L}$ of 1223~\textdegree C and 1231~\textdegree C, respectively. Thus, the application of thermodynamic models to predict the role of high CO$_{2}$ concentrations on phase stability in water-poor systems needs to be reconsidered. This appears to be particularly relevant in Italians volcanic systems, which are often reacting with carbonates. \subsection{Implications for magma dynamics}\label{sec4.2} The data presented above (literature data and our new experimental results) can have profound implications upon the rheological behavior of magmas. In particular, viscosity is highly influenced by the amount of solid phases. This, in turn, might influence the way magma ascends towards the Earth surface and eventually erupts. The Vulcanello peninsula formed by effusions from a 600 m long, 0.5 to 1.5 m wide eruptive fissure (dike) \citep[][and reference therein]{Vetereetal2007}. Based on the analysis of fluid inclusions in quartz xenoliths we assume for the storage of the shoshonitic magma a trapping depth between 21.5 and 17 km (corresponding to the Moho depth) \citep{Zanonetal2003,Peccerilloetal2006}. As such, a possible deep storage reservoir could be hypothesised from which, during the eruptive phase of Vulcanello 1, the magma moved directly to the surface, without intermediate storage into the crust \citep{Davietal2009,Nicotraetal2018}. %\pagebreak Taking into account the water free samples VL46 and VL47 (Figure~\ref{fig5}), having a crystal content of about 20 and 30 area\%, respectively, we can calculate the effective viscosity\break of the crystal bearing system using the models proposed by \citet{Sato2005} and \citet{Vonaetal2011} [${\eta}_{r} = \eta_{\mathrm{eff}}/\eta_{m}$ where $\eta_{r}$ is the relative viscosity, $\eta_{\mathrm{eff}}$ is the effective viscosity of the liquid containing a fraction of crystals, and $\eta_{m}$ is the viscosity of the melt using the model of \citealp{Vetereetal2007}]. Moreover, as required by the \citet{Vonaetal2011} model, using image analysis technique on BSE images collected after experiments, one can extract and characterize the crystal shapes from the input grey-scale slides, without considering the possible presence of nanometric crystals \citep{Cassettaetal2023}. Procedures are described in \citet{DellinoLaVolpe1996} and \citet{Loncaric1998}. The aspect ratio (AS $=$ major axis/minor axis) parameter providing information on the particles elongation was determined following \citep{CoxBudhu2008}. Results indicate that for our experiments the average AS for Pl is 2.5 and for Cpx is 1.7 while Fe--Ti oxides (when present) have AS close to 1. By taking into account these results, viscosity values of 10$^{4.0}$--10$^{4.4}$ Pa${\cdot}$s are obtained for system containing 20 and 30 area\% of crystals, respectively by using the \citet{Vonaetal2011} model. These values are comparable to those obtained using the model proposed by \citet[][i.e. 10$^{3.9}$ and 10$^{4.2}$ Pa${\cdot}$s]{Sato2005}. {Figure~\ref{fig7}} shows the variation of viscosity of the shoshonitic magma as a function of the crystal and volatile contents. The increase in viscosity is clearly related to the decrease of the volatile content and the relative increase of solid phases. \begin{figure} \includegraphics{fig07} \caption{\label{fig7} Variation of viscosity vs. water content for samples reported in Table~\ref{tab3}. Symbols are identical to those reported in Figures~\ref{fig2} and~\ref{fig3} (see text for details).} \end{figure} %\pagebreak Considering that magma ascent in dykes is strongly conditioned by dyke width, {magma} ascent in narrow dykes {could} {be obstructed} by magma crystallization. Critical dyke widths were calculated by \citet{Petfordetal1994} for magma {with felsic composition showing} similar {v}iscosities as those investigated here \citep[see also][]{Scailletetal1996}{. Results} {show that} dykes thinner than 1 m {can hardly} propagate from very large depths. {We are aware that} narrow dykes~ (${<}$1 m) cannot be approximated to a smooth tabular shape {with a constant} dyke width, due to the natural roughness of rock fracture planes. {However, by doing the identical exercise as in \citet{Petfordetal1994}, and taking into account (a) basaltic magma temperature, (b) freezing point, (c) rock country temperature (1200~\textdegree C, 700~\textdegree C and 300~\textdegree C respectively), (d)~the latent heat for basalt (400 J/g) and (e) the s}pecific heat of basaltic magma (1.0 J${\cdot}$g$^{-1}{\cdot}$\textdegree C$^{-1}$), the critical size for dikes reduces to 0.3 m. Thus, we have some constrains for the following discussion. \begin{figure*} \vspace*{-2pt} \includegraphics{fig08} \vspace*{-2pt} \caption{\label{fig8} (a,b) Velocities and ascent times evolution vs. conduits size for Vulcanello (a) and hypothetic volcanic systems with dike widths ${\leq}$3 m and a density contrast between magma and country rock $\Delta\rho=200$ kg/m$^{3}$ (b). Please refer to text for details.} \vspace*{-2pt} \end{figure*} In order to calculate a possible ascent velocity (\textit{u}; m${\cdot}$s$^{-1}$) of these magmas we can use the relation of \citet{ListerKerr1991} for laminar flow: {\begin{equation}\label{eq4} u=(w^{2}/3\eta)\Delta\rho g. \end{equation}}\unskip As stated above, the eruptive fissure at Vulcanello has width ${\leq}$1.5 m. If now we consider a density contrast between magma and country rock of $\Delta\rho=275$~kg/m$^{3}$ as reported in \citet{Vetereetal2007}, and $\eta= 10^{3.5}$ Pa${\cdot}$s for a system containing melt with 0.3~wt\% H$_{2}$O (see also Table~\ref{tab3} samples VL46, VL47), we can estimate a possible rise velocity of such magmas, assuming that the width of the conduits varies from 0.1 to 1.5 m. Results are presented in {Figure~\ref{fig8}a} where the ascent velocity is found to vary in the range \textit{u}(CO$_{2}$ {free}) $=$ 1.5 ${\times}$ 10$^{-4}$--3.5 ${\times}$ 10$^{-2}$ m/s, and \textit{u}(CO$_{2}$ {bearing}) $=$ 5.7 ${\times}$ 10$^{-4}$--1.3 ${\times}$ 10$^{-1}$m/s. The estimated velocity is subject to an abrupt increase if we consider H$_{2}$O--CO$_{2}$ bearing systems. Considering that at Vulcanello the water content is estimated between 0.3 and 1.9 wt\% [analysis of melt inclusions from shoshonitic samples of Vulcanello in \citealp{Gioncadaetal1998} and \citealp{Clocchiattietal1994a,Clocchiattietal1994b}] and considering a temperature of 1080~\textdegree C [the estimated temperature for Vulcanello magmas is 1083 ${\pm}$ 40~\textdegree C, \citealp{Zanonetal2003}], a crystallinity of 17 area\% can be expected for H$_{2}$O- and CO$_{2}$-bearing samples (e.g. VL37; Table~\ref{tab3}). At these conditions the magma viscosity is about 10$^{2.8}$ Pa${\cdot}$s and its ascent velocity (1.5 ${\times}$ 10$^{-2}$--3.3 m/s as $w$ varies from 0.1 to 1.5 m, respectively, Figure~\ref{fig7}) would be faster than in a volatile-free magma {(}see \textit{u}(CO$_{2}$ {free}) given above{)}. Thus, in case of a nominally dry magmatic system rising in conduits having width 1.5 m, the time to reach the surface starting from a depth of 20 km is slightly lower than 4 years. However, if CO$_{2}$ is present, this duration decreases to ${\sim}$300 days, and ${<}$2 h are estimated assuming that 2.0 wt\% H$_{2}$O $+$ 0.6~wt\% CO$_{2}$ are dissolved in the melt. Although these short timescales cannot be applied to the Vulcanello magma, which has lower volatile \citep{Clocchiattietal1994a,Clocchiattietal1994b}, they could be relevant to other shoshonitic volcanic systems. As a further example, by making identical considerations as above, we can consider a hypothetic scenario with a viscosity $\eta = 10^{3.5}$ Pa${\cdot}$s (for a system containing melt $+$ 0.3 wt\% H$_{2}$O see Table~\ref{tab3}, samples VL46, VL47). Thus, if dyke widths are ${\leq}$3 m and the density contrast between magma and country rock is $\Delta\rho=200$ kg/m$^{3}$ and adopting Equation~(\ref{eq4}), we can have an estimate of the possible rise velocity as presented in {Figure~\ref{fig8}b}. In this case, results are: \textit{u}(CO$_{2}$ {free}) $=$ 1.1 ${\times}$ 10$^{-4}$--1.0 ${\times}$ 10$^{-1}$ m/s, and \textit{u}(CO$_{2}$ {bearing}) $=$ 4.2 ${\times}$ 10$^{-4}$--3.7 ${\times}$ 10$^{-1}$ m/s. Finally, if we consider H$_{2}$O--CO$_{2}$ bearing systems with concentrations similar to the experiments VL32 or VL33 (Table~\ref{tab3}) and a temperature of 1080~\textdegree C, the lower crystallinity allows magma to have viscosity in the order of 10$^{2}$ Pa${\cdot}$s and, as consequence, to rise much faster in the range 4.7 ${\times}$ 10$^{-2}$--42.9 m/s as \textit{w} varies from 0.1 to 3 m, respectively (Figure~\ref{fig8}). It is important to keep in mind that in dynamic magmatic systems (during ascent) the CO$_{2}$ release is relatively rapid and possibly can locally accelerate magma ascent. \section{Conclusions}\label{sec5} Crystallization experiments of a natural shoshonite were performed at 500 MPa, and 1080~\textdegree C and water activities \textit{a}H$_{2}$O from 0.01 to 1.00. Results show that clinopyroxene is the main crystallizing phase and plagioclase is found to be a stable phase only for\break water activity lower than 0.1, while spinel is present in all the runs except in those where log \textit{f}O$_{2}$ was lower than ${-}$9 (${\Delta}$NNO ${-}$0.11). Interestingly, at low water activity (below 0.1) the addition of CO$_{2}$ at constant water content reduces the crystallinity of shoshonitic systems. This behaviour is not predicted by classical thermodynamic models. Thus, in the case of CO$_{2}$-rich and H$_{2}$O-poor systems, magmas become less viscous due to the relative decrease in crystallinity. The application of viscosity models to our experimental dataset allows us to infer the possible ascent velocity and ascent time of this type of shoshonitic H$_{2}$O-poor magmas from Vulcanello. Since the magmatic composition used for our study is chemically similar to many Italian volcanic systems (Aeolian Arc, Campi Flegrei, Ischia Island, Pontine Islands, Monti Cimini, Monte Amiata, Capraia Island, Radicofani, Roccamonfina) and also to other systems worldwide (Yellowstone, Mariana Arc, Kurile Arc, Tonga Arc, Andean Arc, Kamchatka Arc), we suggest that our observations will be useful to better understand shoshonitic magma behaviour under relevant geological scenarios. The probability of very rapid ascent of less-evolved melts from depth has to be taken into account for future volcanic crisis and hazard evaluation. \section*{Conflicts of interest} Authors have no conflict of interest to declare. \section*{Acknowledgements} This study was funded by the ``\textit{Piano di Sostegno alla Ricerca 2022 per finanziamenti a progetti di ricerca Curiosity-driven (F-CUR\_CREAMI)''} to FV. \def\circ{\text{\textdegree}} \back{} \bibliographystyle{crgeos} \bibliography{crgeos20230202} \refinput{crgeos20230202-reference.tex} \end{document}