subcrt: Properties of Species and Reactions

Description Usage Arguments Details Value Warning References See Also Examples

View source: R/subcrt.R

Description

Calculate the standard molal thermodynamic properties of one or more species or a reaction between species as a function of temperature and pressure.

Usage

1
2
3
4
5
  subcrt(species, coeff = 1, state = NULL,
    property = c("logK","G","H","S","V","Cp"),
    T = seq(273.15,623.15,25), P = "Psat", grid = NULL, 
    convert = TRUE, exceed.Ttr = FALSE, exceed.rhomin = FALSE,
    logact = NULL, autobalance = TRUE, IS = 0)

Arguments

species

character, name or formula of species, or numeric, rownumber of species in thermo()$OBIGT

coeff

numeric, reaction coefficients on species

state

character, state(s) of species

property

character, property(s) to calculate

T

numeric, temperature(s) of the calculation

P

numeric, pressure(s) of the calculation, or character, Psat

grid

character, type of PxT grid to produce (NULL, the default, means no gridding)

exceed.Ttr

logical, calculate Gibbs energies of mineral phases and other species beyond their transition temperatures?

exceed.rhomin

logical, return properties of species in the HKF model below 0.35 g cm\S-3?

logact

numeric, logarithms of activities of species in reaction

convert

logical, are units of T, P, and energy settable by the user (default) (see T.units)?

autobalance

logical, attempt to automatically balance reaction with basis species?

IS

numeric, ionic strength(s) at which to calculate adjusted molal properties, mol kg^-1

Details

subcrt calculates the standard molal thermodynamic properties of species and reactions as a function of temperature and pressure. For each of the species (a formula or name), optionally identified in a given state, the standard molal thermodynamic properties and equations-of-state parameters are retrieved via info (except for \H2O liquid). The standard molal properties of the species are computed using equations-of-state functions for aqueous species (hkf), crystalline, gas, and liquid species (cgl) and liquid or supercritical \H2O (water).

T and P denote the temperature and pressure conditions for the calculations and should generally be of the same length, unless P is Psat (the default; see water). Argument grid if present can be one of T or P to perform the computation of a TxP or PxT grid. The propertys to be calculated can be one or more of those shown below:

rho Density of water g cm^-3
logK Logarithm of equilibrium constant dimensionless
G Gibbs energy (cal | J) mol^-1
H Enthalpy (cal | J) mol^-1
S Entropy (cal | J) K^-1 mol^-1
V Volume cm^3 mol^-1
Cp Heat capacity (cal | J) K^-1 mol^-1
E Exapansibility cm^3 K^-1
kT Isothermal compressibility cm^3 bar^-1

Note that E and kT can only be calculated for aqueous species and only if the option (thermo()$opt$water) for calculations of properties using water is set to IAPWS. On the other hand, if the water option is SUPCRT (the default), E and kT can be calculated for water but not for aqueous species. (This is not an inherent limitation in either formulation, but it is just a matter of implementation.)

If convert is TRUE (the default), the input values of T and P are interpreted to have the units given by T.units and P.units (default: \degC and bar), and the output values of G, H, S and Cp are based on the units given in E.units (default: calories). If convert is FALSE, the user units (T.units, P.units, and E.units) are ignored, and T and P are taken to be in Kelvin and bar, and the returned values of G, H, S and Cp are in calories.

A chemical reaction is defined if coeff is given. In this mode the standard molal properties of species are summed according to the stoichiometric coefficients, where negative values denote reactants. Reactions that do not conserve elements are permitted; subcrt prints the missing composition needed to balance the reaction and produces a warning but computes anyway. Alternatively, if autobalance is TRUE, the basis species of a system were previously defined, and all elements in the reaction are represented by the basis species, an unbalanced reaction given in the arguments to subcrt will be balanced automatically. The auto balancing doesn't change the reaction coefficients of any species that are do not correspond to basis species.

Minerals with polymorphic transitions (denoted by having states cr (lowest T phase), cr2, cr3 etc.) can be defined generically by cr in the state argument with a character value for species. subcrt in this case simultaneously calculates the requested properties of all the phases of each such mineral (phase species) and, using the values of the transition temperatures calculated from those at P = 1 bar given in the thermodynamic database together with functions of the entropies and volumes of transitions (see dPdTtr), determines the stable phase of the mineral at any grid point and substitutes the properties of this phase at that point for further calculations. If individual phase species of minerals are specified (by cr, cr2 etc. in state), and exceed.Ttr is FALSE (the default), the Gibbs energies of these minerals are assigned values of NA at conditions beyond their transition temperature, where another of the phases is stable. If you set exceed.Ttr to TRUE to calculate the properties of mineral polymorphs (i.e., using cr) the function will identify the stable polymorph using the calculated Gibbs energies of the phase species instead of the tabulated transition temperatures. This is not generally advised, since the computed standard molal properties of a phase species lose their physical meaning beyond the transition temperatures of the phase.

If logact is provided, the chemical affinities of reactions are calculated. logact indicates the logarithms of activities (fugacities for gases) of species in the reaction; if there are fewer values of logact than number of species those values are repeated as necessary. If the reaction was unbalanced to start, the logarithms of activities of any basis species added to the reaction are taken from the current definition of the basis species. Columns appended to the output are logQ for the log10 of the activity product of the reaction, and A for the chemical affinity, in the units set by E.units. Note that affinity provides related functionality but is geared toward the properties of formation reactions of species from the basis species and can be performed in more dimensions. Calculations of chemical affinity in subcrt can be performed for any reaction of interest; however, they are currently limited to constant values of the logarithms of activities of species in the reactions, and hence of logQ, across the computational range.

If IS is set to a single value other than zero, nonideal is used to calculate the adjusted properties (G, H, S and Cp) of charged aqueous species at the given ionic strength. To perform calculations at a single P and T and for multiple values of ionic strength, supply these values in IS. Calculations can also be performed on a P-IS, T-IS or P,T-IS grid. Values of logK of reactions calculated for IS not equal to zero are consistent with the adjusted Gibbs energies of the charged aqueous species.

subcrt is modeled after the functionality of the SUPCRT92 package (Johnson et al., 1992). Certain features of SUPCRT92 are not available here, for example, calculations as a function of density of \H2O instead of pressure, or calculations of temperatures of univariant curves (i.e. for which logK is zero).

For calculations below 273.16 K, the pressure should be set to 1, as \Psat is not defined in these conditions.

If thermo()$opt$varP is TRUE, standard Gibbs energies of gases will be converted from a standard state at 1 bar (as used in SUPCRT) to a variable pressure standard state (see chapter 12 in Anderson and Crerar, 1993). This is useful for constructing e.g. boiling curves for organic compounds.

Value

For subcrt, a list of length two or three. If the properties of a reaction were calculated, the first element of the list (named reaction) contains a dataframe with the reaction parameters; the second element, named out, is a dataframe containing the calculated properties. Otherwise, the properties of species (not reactions) are returned: the first element, named species, contains a dataframe with the species identification; the second element, named out, is itself a list, each element of which is a dataframe of properties for a given species. If minerals with phase transitions are present, a third element (a dataframe) in the list indicates for all such minerals the stable phase at each grid point.

Warning

Although SUPCRT92 prohibits calculations above 350 \degC at \Psat (“beyond range of applicability of aqueous species equations”), CHNOSZ does not impose this limitation, and allows calculations up to the critical temperature (373.917 \degC) at \Psat. Interpret calculations between 350 \degC and the critical temperature at \Psat at your own risk. The discontinuity in the value of \logK at \Psat that is apparent in demo("NaCl") demonstrates one unexpected result.

NAs are produced for calculations at Psat when the temperature exceeds the critical temperature of \H2O. In addition, properties of species using the revised HKF equations are set to NA wherever the density of \H2O < 0.35 g/cm\S3 (threshold just above the critical isochore; Johnson et al., 1992). Both of these situations produce warnings, which are stored in the warnings element of the return value.

NAs are also output if the T, P conditions are otherwise beyond the capabilities of the water equations of state derived from SUPCRT92 (H2O92D.f), but the messages about this are produced by water.SUPCRT92 rather than subcrt.

References

Anderson, G. M. and Crerar, D. A. (1993) Thermodynamics in Geochemistry: The Equilibrium Model, Oxford University Press. https://www.worldcat.org/oclc/803272549

Johnson, J. W., Oelkers, E. H. and Helgeson, H. C. (1992) SUPCRT92: A software package for calculating the standard molal thermodynamic properties of minerals, gases, aqueous species, and reactions from 1 to 5000 bar and 0 to 1000\degC. Comp. Geosci. 18, 899–947. doi: 10.1016/0098-3004(92)90029-Q

Helgeson, H. C., Owens, C. E., Knox, A. M. and Richard, L. (1998) Calculation of the standard molal thermodynamic properties of crystalline, liquid, and gas organic molecules at high temperatures and pressures. Geochim. Cosmochim. Acta 62, 985–1081. doi: 10.1016/S0016-7037(97)00219-6

LaRowe, D. E. and Helgeson, H. C. (2007) Quantifying the energetics of metabolic reactions in diverse biogeochemical systems: electron flow and ATP synthesis. Geobiology 5, 153–168. doi: 10.1111/j.1472-4669.2007.00099.x

Shock, E. L., Oelkers, E. H., Johnson, J. W., Sverjensky, D. A. and Helgeson, H. C. (1992) Calculation of the thermodynamic properties of aqueous species at high pressures and temperatures: Effective electrostatic radii, dissociation constants and standard partial molal properties to 1000 \degC and 5 kbar. J. Chem. Soc. Faraday Trans. 88, 803–826. doi: 10.1039/FT9928800803

See Also

info can be used to find species in the thermodynamic database. makeup is used by subcrt for parsing formulas to check mass balance of reactions. demo("ORP") and nonideal for examples using the IS argument.

Examples

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
## Properties of species
subcrt("water")
# Change temperature
subcrt("water", T = seq(0, 100, 20))
# Change temperature and pressure 	
T <- seq(500, 1000, 100)
P <- seq(5000, 10000, 1000)
subcrt("water", T = T, P = P)
# Temperature-pressure grid
subcrt("water", T = c(500, 1000), P = c(5000, 10000), grid = "P")

## Properties of reactions
subcrt(c("glucose", "ethanol", "CO2"), c(-1, 2, 2), T = 25)
# Use CO2(gas) (or just change "CO2" to "carbon dioxide")
subcrt(c("glucose", "ethanol", "CO2"), c(-1, 2, 2), c("aq", "aq", "gas"), T = 25)

## Automatically balance reactions
# First define the basis species
basis(c("CO2", "H2O", "NH3", "H2S", "O2"))
# Auto-balance adds the required amount of H2O and O2
subcrt(c("ethanol", "glucose"), c(-3, 1), T = 37)
# An example with H+
basis(c("H2O", "H2S", "O2", "H+"))
subcrt(c("HS-", "SO4-2"), c(-1, 1), T = 100)

## Mineral polymorphs
# Properties of the stable polymorph
subcrt("pyrrhotite")
# Properties of one of the polymorphs (metastable at other temperatures)
subcrt(c("pyrrhotite"), state = "cr2")
# Reactions automatically use stable polymorph
subcrt(c("pyrite", "pyrrhotite", "H2O", "H2S", "O2"), c(-1, 1, -1, 1, 0.5))

## Messages about problems with the calculation
# Above the T, P limits for the H2O equations of state
subcrt("alanine", T = c(2250, 2251), P = c(30000, 30001), grid = "T")
# Psat is not defined above the critical point;
# this also gives a warning that is suppressed to keep the examples clean
suppressWarnings(subcrt("alanine", T = seq(0, 5000, by = 1000)))

## minerals with phase transitions
# compare calculated values of heat capacity of iron with
# values from Robie and Hemingway, 1995
T.units("K")
E.units("J")
# we set pressure here otherwise subcrt uses Psat (saturation 
# vapor pressure of H2O above 100 degrees C) which can not be 
# calculated above the critical point of H2O (~647 K)
s <- subcrt("Fe", T=seq(300, 1800, 20), P=1)
plot(s$out[[1]]$T, s$out[[1]]$Cp, type="l",
  xlab=axis.label("T"), ylab=axis.label("Cp"))
# add points from RH95
RH95 <- read.csv(system.file("extdata/cpetc/RH95.csv", package="CHNOSZ"))
points(RH95[,1], RH95[,2])
title(main=paste("Heat capacity of Fe(cr)\n",
  "(points - Robie and Hemingway, 1995)"))
# reset the units to default values
T.units("C")
E.units("cal")

## Subzero (degrees C) calculations
# uncomment the following to try IAPWS95 instead of SUPCRT92
#water("IAPWS95")
# the limit for H2O92D.f (from SUPCRT92) is currently -20 deg C
# but we go to -30 knowing properties will become NA
sb <- subcrt(c("H2O", "Na+"), T=seq(-30, 10), P=1)$out
# start plot with extra room on right
opar <- par(mar=c(5, 4, 4, 4))
# plot Delta G
plot(sb$water$T, sb$water$G, ylim=c(-63000, -56000), xlab=axis.label("T"),
  ylab=axis.label("DG0"))
points(sb$`Na+`$T, sb$`Na+`$G, pch=2)
# add Cp
# change y-axis
par("usr"=c(par("usr")[1:2], -100, 25))
axis(4)
mtext(axis.label("Cp0"), side=4, line=3)
points(sb$water$T, sb$water$Cp, pch=16)
points(sb$`Na+`$T, sb$`Na+`$Cp, pch=17)
legend("topleft", pch=c(16, 1, 17, 2), legend=c("H2O Cp", "H2O G", "Na+ Cp", "Na+ G"))
H2O <- expr.species("H2O")
Na <- expr.species("Na+")
degC <- expr.units("T")
title(main=substitute(H2O~and~Na~to~-20~degC, list(H2O=H2O, Na=Na, degC=degC)))
par(opar)

## Calculations using a variable-pressure standard state
thermo("opt$varP" = TRUE)
# Calculate the boiling point of octane at 2 and 20 bar
# We need exceed.Ttr=TRUE because the liquid is metastable
# at high temperatures (also, the gas is metastable at low
# temperatures, but that doesn't produce NA in the output)
sout2 <- subcrt(rep("octane", 2), c("liq", "gas"),
  c(-1, 1), T = seq(-50, 300, 0.1), P = 2, exceed.Ttr = TRUE)$out
sout20 <- subcrt(rep("octane", 2), c("liq", "gas"),
  c(-1, 1), T = seq(-50, 300, 0.1), P = 20, exceed.Ttr = TRUE)$out
# find T with the Gibbs energy of reaction that is closest to zero
Tvap2 <- sout2$T[which.min(abs(sout2$G))]
Tvap20 <- sout20$T[which.min(abs(sout20$G))]
# Compare these with experimental values (Fig. 1 of Helgeson et al., 1998)
Tvap2.exp <- 156
Tvap20.exp <- 276
# Reset varP to FALSE (the default)
thermo("opt$varP" = FALSE)

CHNOSZ documentation built on April 9, 2021, 9:08 a.m.