Geochemical Modeling - Science topic

check on geochemist workbench from the link here https://www.gwb.com/

Good luck.

Use R-package : psych

Tuan Quang Tran Thank you for your valuable suggestion and time.

Regards

Divyadeep harbola

_{I suggest you open an account on : https://phreeqcusers.org/index.php?action=login2 Follow forum sections You will find a lot of educational resources} Best

Hi Sabina,

Not easy to compare as each one is dedicated to different applications.

PHREEQC is widely popular among the geochemical modeling community and is considered by far the most complete geochemical modeling package. It is very flexible and easy to use. In version 3, there is support for modeling high-pressure-temperature systems.

TOUGHREACT couples a geochemical engine (REACT) with the well-known multiphase and multicomponent TOUGH2 code (developed by LBNL) which has well-known applications in geothermal reservoir engineering and CO2 storage applications.

The code REACT has fewer capabilities for pure geochemical modeling compared to PHREEQC, and to my best knowledge, it is not (or much less) maintained than PHREEQC. Hence, the latter is better for pure geochemical models or 1D fully saturated reactive transport simulations.

TOUGHREACT is advised for multi-dimensional (2D or 3D) reactive transport applications possibly with multiphase fluid flow and transport.

Hope this can help you.

Dear colleague,

I would suggest R/GCDkit. We have published a book on numerical modelling of igneous petrogenesis a few years back:

Janoušek, V., Moyen, J. F., Martin, H., Erban, V. & Farrow, C. M. (2016). Geochemical Modelling of Igneous Processes – Principles and Recipes in R Language. Bringing the Power of R to a Geochemical Community. Springer-Verlag, Berlin, Heidelberg, 346 pp. doi:10.1007/978-3-662-46792-3

I trust that this helps,

Vojtech Janousek

PS

For further info, see:

some freebies are here: http://book.gcdkit.org/

######################################################################### # Run ex2 and plot results ######################################################################### # load the phreeqc.dat database phrLoadDatabaseString(phreeqc.dat) # run example 2 phrRunString(ex2) # retrieve selected_output as a list of data.frame so <- phrGetSelectedOutput() # plot the results attach(so$n1) title <- "Gypsum-Anhydrite Stability" xlabel <- "Temperature, in degrees celcius" ylabel <- "Saturation index" plot(temp.C., si_gypsum, main = title, xlab = xlabel, ylab = ylabel, col = "darkred", xlim = c(25, 75), ylim = c(-0.4, 0.0)) points(temp.C., si_anhydrite, col = "darkgreen") legend("bottomright", c("Gypsum", "Anhydrite"), col = c("darkred", "darkgreen"), pch = c(1, 1))

How can I calculate pCO2 in groundwater using by Model PHREEQC

Kgaugelo Thobejane

First of all, download LLNL database from the source of the software¹. Then add the mineral properties you interested into the window . You see a snapshot of the editing page of the thermodynamic database in the file added.

¹ https://www.gwb.com/thermo.php#TEXT

It is an iterative calculation

Look at here

http://chemequ.ru/%D1%81%D0%BF%D1%80%D0%B0%D0%B2%D0%BE%D1%87%D0%BD%D0%B8%D0%BA%D0%B8/%D0%BA%D0%BE%D0%BD%D1%81%D1%82%D0%B0%D0%BD%D1%82%D1%8B_%D0%BA%D0%BE%D0%BC%D0%BF%D0%BB%D0%B5%D0%BA%D1%81%D0%BE%D0%BE%D0%B1%D1%80%D0%B0%D0%B7%D0%BE%D0%B2%D0%B0%D0%BD%D0%B8%D1%8F/

Good luck

Dear Professor

The materials are sourced from an Andesite quarry.

Samples are oriented on 2.54 cm slides that were prepared after three stages of pre-treatment procedure to isolate clay-sized particles. I presented the XRD results on air-dried samples. I also have the XRD test results for sample which are glycolated at 60 degree of Celsius for 20 hours and the heated sample as well (45 minutes at 550 degree of Celsius). I can provide them in the case of need.

Experiments were run using a PANalytical Empyrean X-ray diffractometer operating at 45 kV and 40 mA. The diffractometer employs a Cu K-alpha radiation (Kα1/Kα2 = 0.5) with a 1.5406 Å wavelength (Kα1) and has a graphite monochromator with a PIXcel Detector. Fresh bulk materials scans were run at a 2θ angle from 4 to 70°, with a 0.013° step size and an integrated 100 s dwell time.

Thank you for your time.

Dear Sir

I think the SHPECK is the suitable software for your research. For more information, please visit the following link

Best Regards

Thanks a lot.

Iron oxides, iron hydroxides, and iron oxy-hydoxides can all be those colors you describe, as well as they are some of the most common. Depending on the oxidation state and crystal arrangement colors can vary from silver to brown to red to orange to yellow. These minerals include things such as the minerals hematite, goethite, and limonite. Common mineral associations can be in the sulfur realm of things, with elements such as pyrite, which is iron sulfide, and other sulfate-reducing mineralizations.

Hi Christian

I recommend you read these articles:

Estimating groundwater mixing ratios and their uncertainties using a statistical multi parameter approach. Journal of Hydrology 305 (2005) 1–14

A methodology to compute mixing ratios with uncertain end-members. WATER RESOURCES RESEARCH, VOL. 40, W12101, doi:10.1029/2003WR002263, 2004

Mixing of groundwaters with uncertain end-members: case study in the Tepalcingo-Axochiapan aquifer, Mexico. Hydrogeology Journal (2012) 20: 605–613 

Regards

For partial melting, it depends on the source. An amphibolite source in the deep crust would have a garnet-bearing residuum that would be depleted in plagioclase feldspar, so you might expect, for example high La/Yb and Sr/Y in the melt (the so-called adakite signature). At lowqwer pressure (less than 8 kb or so) garnet won't be stable and you will lose the garnet signature. If you have a biotite gneiss for a source, garnet will also be stable (even mid-crustal pressure) and the melt will be strongly enriched in LILE elements. I would look for these types of signatures before you start any kind of melting model. I would also point out that most granitoids, even S-types,  are not going to be simple partial melts, but complex mixtures of material derived from both crust and mantle.

You can also check the following

Srivastava, Rajesh K., and Anup K. Sinha. "Geochemistry and petrogenesis of early Cretaceous sub-alkaline mafic dykes from Swangkre-Rongmil, East Garo Hills, Shillong plateau, northeast India." Journal of Earth System Science 113.4 (2004): 683-697.

Rao, J. Mallikharjuna. "Petrology and Geochemistry of Dolerite Dykes, West Gar0 Hills, Meghalaya: A Preliminary Study." Gondwana Research 5.4 (2002): 884-888.

Choudhury, D. K., et al. "Geochemistry and petrogenesis of anorogenic (?) granitoids of west Garo Hills, Meghalaya." Journal of the Geological Society of India 80.2 (2012): 276-286.

Hi! There are different spatial and temporal considerations here. There are also sampling aspects to consider.

Regarding spatial considerations - generally locations for background concentrations are those furthest away/most upstream (or upcurrent) from any known PAH sources. If you have data from the past 10 years, the areas that consistently had the lowest concentrations are a good place to start! These are likely the background areas. These can be further confirmed depending on other data (e.g. if you have current models, are these areas unaffected by currents originting from known PAH sources). Keep in mind that many legacy PAH sources can exist, so good to have many places if you are not sure.

Regarding temporal - in general PAH atmospheric concentrations (both particle and gaseous) are known to increase in the winter due to increase fuel burning (coal, wood, diesel, etc.). So many air monitoring studies havenoticed a winter spike; particularly in valleys or downriver from industrial areas (this should be not hard to find studies on this). So if you are going for yearly trends; sampling during each of the four seasons is recommended as a minimum. Otherwise, you get the lowest concentrations in summer.

Regarding sampling: if you just analyse water samples directly, you will only get a snap shot in time. If you take a passive sample deployed over several months, you can get a time-integrated sample. The passive sampling method that is most popular for PAHs in water is thin polyethylene strips spiked with performance reference compounds (PRCs). In my experience these should be deployed for about three months to get good, time-integrated data; but some deploy these for less (1 month, sometime two weeks). Researchers like Rainer Lohmann and Phill Gschwend have used these extensively. Note also that water sample measurements are for the dissolved and particle phase; whereas passive samples just due the dissolved phase, which more resembles bioavailability. My lab is also developing passive sample techniques that are robust. I made a film on this https://www.youtube.com/watch?v=xZnQt0IKlRE

Regarding sampling frequency: n = 3 per time/location if you have little temporal resolution. n = 2 if you have lots of samples (but take n = 3 at some places). 

Tips: sample water in amber glass bottles (PAHs are light sensitive); and either freeze or add a biocide (like NaN3) after analysis to prevent microbial degradation. 

Other suggestions:

- deploy sediment traps to get the particle phase of PAHs that settle over time in the settling particles.

- Try to seperate particle and dissolve phases while you sample

-Confirming background concentrations in the sediment is also usefull, as well as through air passive sampling, is also useful.

Good luck! :-)

, 

Thank-you Dr Igwe for taking the time to answer this question. While I do understand the need for this type of normalization procedure to harmonize models of sorption derived using different underlying assumptions of surface site density, my problem relates to that it does not seem to work for the surface protonation reactions. The change in delta_pKa implicit in the pKa normalization procedure fundamentally changes the shape of the sorption edge so that a match is not possible, irrespective of the specific surface area or fluid volume/sorbent mass ratio used in the calculation. This, at least appears to be the case when making forward calculations of models described in the literature at different site densities using PhreeqC, which I have now also replicated in VMinteq so it doesn't seem to be a code related issue. Basically, normalization only seems to "work" if it is applied only to the solute binding reactions and not the surface charging (protonation) model. This is at odds with the NEA sorption project recommendations (& the analyses by Kulik, Sverjensky, and others), thus my confusion.

Mr. Sangsefidi

Ion exchange usually yields secondary calcite during Na-activation of bentonites (smectite rich clays) at industrial scale. The same process may easily occur in nature when Na-rich pore waters flow through Ca-bentonites in saline/alkaline environments. Similarly ion exchange between smectites and zeolites may retain K in the system which during diagenesis may facilitate illitization (see the attached paper). On the other hand one should not confuse ion exchange with specific adsorption due to fundamental differences (stoichiometric vs non-stoiciometric process).

On the other hand, multi-valence cations such as REE do not easily exchange in clay minerals, but may be specifically adsorbed on the clay edges at non-permanent charge sites especially at relatively low pH below the pzc of the edges. Bearing in mind that REE are usually bound in trace heavy minerals, sometimes reported to have formed diagenetically by authigenic processes (e.g. monazite-see the attached paper), it is expected that specific adsorption of REE of clay minerals may restrain authigenesis of such REE minerals.

Regards

G. E. Christidis

The subduction-related fluid are riched by LIL elements (Sr, Pb, Rb, K). If the basalts are high-potassium, then the magma protholite contains phlogopite. A positive correlation of K/La to K in basalts is an indicator of the presence of phlogopite and water in the mantle. Restitic association with melting hydrated mantle comprises a titanium-containing phase (e.g., rutile) - Ta-Nb minimum in melts.

I have one paper, which is attached. I would review the PHREEQC webpages http://wwwbrr.cr.usgs.gov/projects/GWC_coupled/phreeqc/. I think they have mailarchive and FAQ that have some info on prior applications.

Also, the other USGS model for surface water reactive transport is OTIS http://water.usgs.gov/software/OTIS/ and there may be some way to couple them.

Other models coupled with PHREEQC include PHT3D http://www.pht3d.org/pht3d_about.html, PHAST http://wwwbrr.cr.usgs.gov/projects/GWC_coupled/phast/, and HP1 http://www.pc-progress.com/en/Default.aspx?h1d-hp1 , but these are generally for groundwater.

Not an expert here at all.

Are you asking for something like this?

Or to just generally reduce hygroscopic nature of the compounds.  The latter I can only say keep them aware from moisture and keep it cold

I agree with Pieter Bertier that the book of Appelo & Postma is an excellent book for people working in hydrochemistry.

About your question I give an example of selected output of one simulation and some comments:

----------------------------Description of solution-------------

                                        pH  =   7.200   

                                       pe  =   8.400   

       Specific Conductance (uS/cm, 17 oC) = 523

                          Density (g/cm3)  =   0.99903

                        Activity of water  =   1.000

                           Ionic strength  =  8.293e-003

                       Mass of water (kg)  =  1.000e+000

                    Total carbon (mol/kg)  =  2.430e-003

                       Total CO2 (mol/kg)  =  2.430e-003

                      Temperature (deg C)  =  17.300

                  Electrical balance (eq)  =  8.449e-005

 Percent error, 100*(Cat-|An|)/(Cat+|An|)  =   0.73

(Comment: This value shows that the % balance error is small (less than 1%). As a rule of thumb analysis with an error <5% are accepted, but this depends on other factors.)

Iterations  =   8

                                  Total H  = 1.110153e+002

                                  Total O  = 5.551980e+001

------------------------------Saturation indices-------------------------------

     Phase               SI log IAP  log KT

     Anhydrite        -2.11   -6.44   -4.34  CaSO4

     Aragonite        -0.69   -8.98   -8.29  CaCO3

     Calcite          -0.54   -8.98   -8.44  CaCO3

     Chalcedony       -0.08   -3.72   -3.64  SiO2

     Chrysotile       -7.53   25.65   33.18  Mg3Si2O5(OH)4

     CO2(g)           -2.15   -3.52   -1.37  CO2

     Dolom_desord     -1.93  -18.27  -16.34  CaMg(CO3)2

     Dolomite         -1.36  -18.27  -16.91  CaMg(CO3)2

     Fluorite         -2.27  -12.96  -10.69  CaF2

     Gypsum           -1.86   -6.44   -4.58  CaSO4:2H2O

     H2(g)           -31.20  -34.32   -3.12  H2

     H2O(g)           -1.71   -0.00    1.71  H2O

     Halite           -7.18   -5.61    1.56  NaCl

     Melanterite      -7.46   -9.77   -2.31  FeSO4:7H2O

     O2(g)           -23.47  -26.30   -2.83  O2

     Quartz            0.38   -3.72   -4.10  SiO2

     Sepiolite        -5.07   10.90   15.97  Mg2Si3O7.5OH:3H2O

     Sepiolite(d)     -7.76   10.90   18.66  Mg2Si3O7.5OH:3H2O

     Siderite         -1.47  -12.31  -10.84  FeCO3

     SiO2(a)          -0.94   -3.72   -2.78  SiO2

     Talc             -4.09   18.21   22.30  Mg3Si4O10(OH)2

Comment: In this example the water is undersaturated with respect of all phases because the saturation indices are always  <0. The solution is close to the equilibrium with respect to Chalcedony.

The CO2 SI shows the partial pressure of CO2 in equilibrium with the solution, in this example  log pCO2 = ‑2.15 which is a value normal for the CO2 pressure in soils in temperate climate.

In general the databases do not include the dolomite desord so this phase was introduced in the input: PHASES

                    Dolom_desord

                    CaMg(CO3)2 = Ca+2 + Mg+2 + 2 CO3-2

                    log_k = -16.58

                    analytical_expression -38.539 0.0217 4618  0  0 

If the solution contains Fe and we don’t introduce the redox state (v.g. pe or some redox pair that permits the calculation of that state, the saturation indices about species including iron are meaningless. This comment applies to all species and phases that are redox state dependent. 

I hope that this answer will be helpful

Dear feng

to have a temperature estimation (but always make attention to the results that you get) you can use these two geothermometers:

- Putirka, K., 2008. Thermometers and Barometers for Volcanic Systems. In: Putirka, K., Tepley, F. (eds.). Minerals, Inclusions and Volcanic Processes: Reviews in Mineralogy and Geochemistry. Mineralogical Society of America, 69, 61-120.

- Thy, P., Lesher, C.E., Tegner, C., 2013. Further work on experimental plagioclase equilibria and the Skaergaard liquidus temperature. American Mineralogist, 98, 1360-1367.

The first one is a book where you can find,  among the other, a geothermometer based on the cpx-liquid equilibrium. Prof. Putirka in his personal website provides the excel file with the algorithms (http://www.fresnostate.edu/csm/ees/faculty-staff/putirka.html).

The second geothermometer works using the equilibrium plagioclase-liquid 

as you can imagine from the title of the paper. 

Cheers, Fabio

Dear Reza

MMI geochemistry is proven for mineral exploration (hidden deposits) by more than 1000 case studies. I have attached my article on MMI geochemistry (in Persian). 

Best Regards

Morteza

Dear  Md.Badriadib,

1000 L = 1 m³

1 mg/Lit    = 0.001 mg/m³

1 mg/L = ppm

Relation of Mass & Volume 22.4 Nm³= 1 Kg. mole (STP,)

                                                 22.4 lit = 1 gram mole(STP)

                                                 385 ft³ = 1 lb mol (STP)

Regards,

Prem Baboo

Dear Ebrahim Sangsefidi

Soil physical, chemical and to some extent biological properties are inter-linked. It will be correct even to say that interactions among two properties definitely can alter the third one. So it is yet a matter of investigation to assess what chemical reactions could affect the soil strength and vice versa.

Hello Craig,

I am afraid that you will need to be more specific in your question.

What conditions are you looking at for a start? Are you looking at dissolving UO2 up to equilibrium (from a mineralogical assemblage, from pure UO2) at a given pH? Or are you looking to dissolve UO2 as a function of time using a kinetic expression (dissolution equation)?

Think about it as a lab experiment, the basic steps would be to:

1/ Generate the solution you want your UO2 to dissolve into

2/ Add the UO2 (or any other minerals)

3/ wait for the reaction to happen

Within PHREEQC a simple exemple would be:

SOLUTION 1

temp 25

pH 5

Na 0.1

Cl 0.1 charge

EQUILIBRIUM_PHASES 1

Uraninite 0 1

END

This calculation will give me the amount of uraninite that can dissolve (SI=0 from 1 mole added) at equilibrium in a pH 5 solution for exemple. 

If you provide more information about what you want to do, a better model can be devised.

Dear Kai Fauth,

Thank you for your response. I will certainly go through your suggestions.

If you wish to purify magnesium ions in a set solution I'd rather look into different separation processes (e.g. chromatography or liquid-liquid extractions) rather than using acids and bases to precipitate inorganic cations and anions from solution. This is also inherently difficult to simulate with software like PHREEQC, because of the formation of solid solutions: when you would precipitate calcium carbonate, you'd inherently also remove magnesium from solution due to the formation of magnesian calcite.

But if you're set at precipitating calcium sulfate and carbonate, i wouldn't look at changing the pH i would use different ions that are relatively insoluble with set ion. Such as barium for the precipitation of sulfate (although you'd have to look into the kinetics as well, because barium sulfate might be kinetically unfavourable to precipitate). 

TOC is a problematic value.  It will be the sum of all losses in weight resulting from decomposition of organic matter, hydromicas and carbonates in the sample.  You may use it as a "quick and dirty method" but unless you know the mineral composition of the\actual deposit, your results will not necessarily be a true measure of the hydrocarbon content of the deposit.  Remember too, that sediments usually vary in composition laterally and this must be taken into account when interpreting the results.  This means that unless you have a substantial amount of data from the same stratum in a relatively small area, your results need to be interpreted with caution..

 Thanks Laurance for the references. I think the fractionation occurred at low temperatures are due to equilibrium fractionation mechanism rather than diffusion, whether during formation of secondary minerals or surface adsorption.       

There may be no specific reason why igneous rocks were used and metamorphic rocks were not used, and it may be that the application of the research was for an area with igneous rocks (without metamorphic rocks). I do not believe there are any specific water-rock interaction differences between igneous and metamorphic rocks. A metamorphic rock is a type of rock which has been changed by extreme heat and pressure. So, an igneous rock can become metamorphic by exposing it to heat and pressure. This heat and pressure typically dissolves and re-precipitates some of the minerals in the rock. So, there are some differences in types and chemistry of the minerals, and the morphology of the minerals may be different, which may impact dissolution. Hope that helps.

Dear Dr. Liu,

Many thanks for the second image.

During the cooling of a basaltic or granitic magma, the anorthite-rich plagioclase crystallize first. The adjoining plagioclase zones are increasingly albite-rich (Na), since the anorthite component (Ca) was incorporated into the inner (early) plagioclase zones.

A subsequent (low temperature) deuteric alteration accesses the anorthite-rich zones at first because anorthite is only stable at high temperatures and unstable at low temperatures. The albite-rich zones, however, are hardly altered since they are stable at low temperatures. The water rich solutions separated from the same magmatic body and reacted with the primary magmatic minerals. Since the hydrothermal fluids can easily penetrate into the innermost regions of the plagioclase, the anorthite-rich core zones are selectively replaced by secondary minerals.

Best regards,

G. Grundmann

For further reading, you can check the attached file.

Dear Feng

if you have the trace element concentrations of clinopyroxene, you can calculate the trace elements composition of the parental liquid using the experimentally determined partition coefficients of Ionov et al. 2002 Journal of Petrology, 43, 2219-2259. 

Best regards 

When reaction is intensive and considerable absorption or dissolution take place within the time frame of the studied problem, it's require to take into account geometrical properties change e.g. change in reactive surface area, change in permeability and porosity, by solving coupled solid mass conservation equation in addition to the ADRE and fluid conservation equation.

please read this paper.

I suggest that you contact the United Nations University Geothermal Training Program in Iceland for help. Good luck!

Late Devonian basaltic volcanism and its possible significance for the F-F boundary was discussed in Mahmudy Gharaie et al. 2004, Geological Querterly, 48 (4): 323-332.  Access to the paper is free on the internet.

Hello Daniel,

a large variety of geochemical modelling codes exist. I can offer you a small choice besides PHREEQC: The Geochemist's Workbench (http://www.gwb.com/), EQ3/EQ6 (https://missions.llnl.gov/energy/technologies/geochemistry), VisualMinteq v.3.1 (http://www2.lwr.kth.se/English/OurSoftware/Vminteq/) and Wateq4f (http://wwwbrr.cr.usgs.gov/projects/GWC_chemtherm/software.htm).

I am not sure If one of this codes provides a COM interface.

You can find lists of codes here:

Hope that helps!

Dear Dr. Mohammadzadeh,

I have been dealing with this matter since some years in slightly different way.

Maybe that this paper may help you. If you send me an E mail via RG, I will forward to you a copy privately:

Looking forward to hear from you.

Best regards

Harald G. Dill

You may use PHREEQ model for understanding dissolution-precipitation process in groundwater. Based on the groundwater quality data at upstream and downstream points along a pathline, geochemistry of biotite mineral can be studied. You may refer following papers where precipitation-dissolution of calcite, fluorite and other minerals are studied:

1. S.D. Dhiman and Ashok K. Keshari (2006) GIS assisted inverse geochemical modeling for plausible phase transfers in aquifers. Environmental Geology, 50(8): 1211-1219.

2. S.D. Dhiman and Ashok K. Keshari (2006) Hydrogeochemical evaluation of high fluoride groundwaters: A case study from Mehsana district of Gujarat, India. Hydrological Sciences Journal, 51(6), 1149-1162.

In multivariate analysis, an interesting situation arises when we have , e.g., K element in ground water and we normalize data before calculating correlation matrix, it may be bimodal distribution and we won't be able to normalize the data. What do we do then?

Phreeplot  for your purpose. Try it  with  Phreeqc or PhreeqcI

Dear Gautam

Is IRMS laboratory available over there in Hyderabad?

Do you want to use your geochemical data (major- and trace elements) in order to find out whether or not you're dealing with FC-AFC-FCA or magma mixing processes? If so i think Ersoy & Helvaci (2010) published a paper that might be helpful, since it also contains an excel spreadsheet in the supplementary material.

Bonjour,

You can find on Researcher Gate the basics for establishing the different melting models in the first part of my thesis: Bougault, H. (1980). Contribution des éléments de transition à la compréhension de la génèse des basaltes océaniques. Analyse des éléments-traces dans les roches par spectrométrie de fluorescence X. Paris, Paris VII, France: 221.

Long time later, I pursued this work through a student, Jean Marc Commenge: this work adds the dynamic of upwelling ma,tle beneath the Ridge and the relative upwelling speeds of the solid and liquid material.

I just put this material on Researcher Gate and find it enclosed.

In addition I have the XL files that were used for computing and generate figures that I will put as well on Researche Gate.

Henri Bougault

Hello Shanmuga,

I think both approaches – as suggested by Lukasz and David – are necessary. In addition, most important is the knowledge of the local geogenic background concentrations of the estuarine environment. In the case of metals in fluviatile sediments these values correspond to the stable fraction portion. This is reasonable because during fluviatile transport processes only the very stable minerals reach the estuarine region. However, during transport additional compounds of anthropogenic provenience (industry, urbanic, agriculture) may be added to the sediment during transport. These compounds correspond to the environmental impact and can be identified by chemical speciation analysis (sequential elution). During the sequential elution procedure the contaminants of fluviatile sediments are stripped of step by step. The last step yields the stable fraction, which in general corresponds to the uncontaminated background composition of metals.

We published on this problem in:

“Floodplain lakes as an archive for the metal pollution in the River Elbe (Germany) during the 20th century“ 

D. W. Zachmann et al.; Geochemistry 06/2013; 14 – 27.

“Assessment of element distribution and heavy metal contamination in Chilika Lake sediments (India)” D. W. Zachmann et al.; Lakes & Reservoirs Research & Management 06/2009; 14(2):105 - 125.

“Hintergrundwerte und Schwermetallbelastung von Sedimenten im Bereich von Fließgewässern am Beispiel des Oderhaffs” D. W. Zachmann et al.; Mitt Umweltchem Oekotox. 03/2008; 14(Jahrg.(1)):11-14.

Best wishes and kind regards

D. W. Zachmann

Technically there is a direct relationship based on the the equilibrium between oxygen and water. You could rearrange the relationship K=aH2O/P1/2O2 x a2e- x a2H+ such that for any pH and O2 values you can calculate ae- and thus convert pe to Eh. The equation is in in most geochemical texts. As mentioned above this assumes that you don't have any other redox couples in the system. In practical terms, that assumption is usually not met. However, except for extreme cases the oxygen dominates Eh in natural waters until the dissolved O2 is less than 2 ppm. You can also as suggested above use PHREEQC, etc. to calculate the Eh, but again the function is relatively insensitive and the calculated Eh will remain very high until the O2 falls below about 2 ppm.

You might wish to consult the paper:

The difference of diffusion coefficients in water for arsenic compounds at various pH and its dominant factors implied by molecular simulations.

Masato Tanakaa, Yoshio Takahashia, Noriko Yamaguchic, Kyoung-Woong Kimd, Guodong Zhengb, Mika Sakamitsua.

Geochimica et Cosmochimica Acta, Volume 105, 15 March 2013, Pages 360–371

And the references therein.

There could be various approaches to determine such, either from direct analysis of mantle xenoliths or via inverse modelling from a dataset on primitive mantle-derived melts. As a result you will see that the composition of upper mantle can vary significantly. If you consider that mantle is not a homogeneous source, but rather a mixture of different lithologies the answer will be still more complicated. In addition to the reference given to you by Chao Zhang above I can suggest another reference such as Pilet S., Baker M.B., Stolper E.M. Metasomatized lithosphere and the origin of alkaline lavas // Science, 2008, v. 320, p. 916-919.

If you read Russian, you can go and see my habilitation thesis which includes some discussion on such a topic.

In Chapter 1 https://dl.dropbox.com/u/74801948/018-079-Chapter-1.pdf you will find Table 1.1 with major element compositions of variable sources such as depleted MORB mantle (DMM), fertile peridotite PHN1611, pyrolite (R1975, MS1995), piclogite (66SAL-1, MIX1G) and eclogite (Gb108, G2, Нкк). Trace element compositions are available for some of these compositions in Workman, Hart, 2005 (DMM), McDonough, Sun, 1995 (MS1995), Yaxley,

Sobolev, 2007 (Gb108), Rudnick, Fountain, 1995 (Нкк which is Russian for LCC). In Chapter 3 I performed some simple modelling where I used as a source of mantle either DMM or an arbitrary mix between DMM and various types of eclogite-piclogite with trace element composition of MORB or real compositions from literature. An example of such simple modelling you can find elsewhere in literature. I give reference to my paper (Ivanov A.V. Evaluation of different models for the origin of the Siberian traps. / Foulger G.R., Jurdy D.M., eds. The origin of melting anomalies: Plates, plumes and planetary processes // Geological Society of America Special Paper, 2007, v. 430, p. 669—691. Fig. 10, for La and Yb http://www.academia.edu/492738/Evaluation_of_different_models_for_the_origin_of_the_Siberian_Traps) simply because it is the easiest way to me to pick up a reference. But you can find much more in the literature.

Hope it helps.

For another example, try: Beier et al 2006 "Magma Evolution of the Sete Cidades Volcano, Sao Miguel, Azores JOURNAL OF PETROLOGY VOLUME 47 NUMBER 7 PAGES 1375–1411 2006 doi:10.1093/petrology/egl014

Regards

Graham Howes