Potassium versus Sodium in Sea Water

[WillyEngland]

Why is there so much more sodium than potassium in sea water?

ca. 30 : 1.

[exchemist]

4 hours ago, WillyEngland said:

Why is there so much more sodium than potassium in sea water?

ca. 30 : 1.

Interesting question. I see the abundance of K in the earth's crust  is similar to that of Na: http://hyperphysics.phy-astr.gsu.edu/hbase/Tables/elabund.html . Both are of course alkali metals with many soluble salts.

I have a feeling it may be to do with the stability of complex minerals containing K, e.g. aluminosilicates. Perhaps the greater ionic radius of K+ forms more energetically favourable lattices , or it is harder for the larger ion to migrate within minerals and eventually be leached out by water. But I confess I am guessing. I would be interested to know the reason. Perhaps someone more knowledgeable can comment.

Edited January 19 by exchemist

[swansont]

52 minutes ago, exchemist said:

I see the earth abundance of K  is similar to Na

It’s 2,000,000 ppb for Na and 110,000 ppb for K, which might account for a large part of this

https://en.m.wikipedia.org/wiki/Abundance_of_the_chemical_elements

[sethoflagos]

1 hour ago, exchemist said:

I have a feeling it may be to do with the stability of complex minerals containing K, e.g. aluminosilicates. Perhaps the greater ionic radius of K+ forms more energetically favourable lattices , or it is harder for the larger ion to migrate within minerals and eventually be leached out by water.

Reckon so. I've seen somewhere that albite (sodium feldspar) weathers at 10 times the rate of orthoclase (potassium feldspar). 

[exchemist]

14 minutes ago, swansont said:

It’s 2,000,000 ppb for Na and 110,000 ppb for K, which might account for a large part of this

https://en.m.wikipedia.org/wiki/Abundance_of_the_chemical_elements

Well those numbers are certainly different from the ones I had found. But even so, if you look further down in the article, at the graph for abundance in the upper crust, which is where the minerals in the sea would be leached from, the difference seems to be considerably less.  Here's another source also suggesting levels of the two in the crust are comparable: https://pressbooks.lib.vt.edu/introearthscience/chapter/3-minerals/. 

 

 

 

 

2 minutes ago, sethoflagos said:

Reckon so. I've seen somewhere that albite (sodium feldspar) weathers at 10 times the rate of orthoclase (potassium feldspar). 

Ah. Maybe I'll have a look into that later - I need to go and cook a kedgeree for supper. 

[swansont]

1 hour ago, exchemist said:

Well those numbers are certainly different from the ones I had found. But even so, if you look further down in the article, at the graph for abundance in the upper crust, which is where the minerals in the sea would be leached from, the difference seems to be considerably less.  Here's another source also suggesting levels of the two in the crust are comparable: https://pressbooks.lib.vt.edu/introearthscience/chapter/3-minerals/. 

 

Comparable, but also that’s by weight, so Na is ~ 2x more abundant by number.

[exchemist]

1 hour ago, swansont said:

Comparable, but also that’s by weight, so Na is ~ 2x more abundant by number.

Aha, that makes a difference, certainly!

But there is a factor of 30 to account for.

Edited January 19 by exchemist

[KJW]

46 minutes ago, exchemist said:

But there is a factor of 30 to account for.

It could be that the earth's crust has already been depleted of its sodium by all the rainfall.

 

[exchemist]

25 minutes ago, KJW said:

It could be that the earth's crust has already been depleted of its sodium by all the rainfall.

 

Yeah but the point is why would rainfall deplete Na more than K by a factor of 30, given that there are similar amounts - within a factor of 2 or so - in the Earth's crust to start with, and both form equally soluble cations, more or less. 

[KJW]

17 minutes ago, exchemist said:

Yeah but the point is why would rainfall deplete Na more than K by a factor of 30, given that there are similar amounts - within a factor of 2 or so - in the Earth's crust to start with, and both form equally soluble cations, more or less. 

But I'm suggesting that there was about 30 times more sodium than potassium in the earth's crust to begin with, and that over time this became depleted to the ratio we see today. In that case, why does sodium dissolve preferentially? Maybe sodium was in a more soluble form (eg chloride) than potassium (eg something else... I don't know what). While both sodium and potassium tend to form soluble salts, I think potassium is more likely to form a less soluble salt.

 

[exchemist]

29 minutes ago, KJW said:

But I'm suggesting that there was about 30 times more sodium than potassium in the earth's crust to begin with, and that over time this became depleted to the ratio we see today. In that case, why does sodium dissolve preferentially? Maybe sodium was in a more soluble form (eg chloride) than potassium (eg something else... I don't know what). While both sodium and potassium tend to form soluble salts, I think potassium is more likely to form a less soluble salt.

 

The latter point is what I too suspect, to do with ion sizes. But I’m not a mineralogist. 
 

We need to check what these abundance numbers mean. Do they include the oceans or not? 

[TheVat]

I was taught (in a geology course) that sodium cation is the least stable in water, so it tends to more quickly hook up with a chloride or other anion, and so becomes the dominant salt.  That, paired with sodium being the most easily leached from surface rocks as rain washes over, and you get saline oceans.  In fact, the oceans would be even more saline were it not for subduction.  And sedimentation.

Was tempted to add, take my comment with a grain of salt.

 

Edited January 19 by TheVat

[sethoflagos]

1 minute ago, exchemist said:

The latter point is what I too suspect, to do with ion sizes. But I’m not a mineralogist. 
 

We need to check what these abundance numbers mean. Do they include the oceans or not?

If this helps, the primary route for transport of sodium and potassium into the earth's crust from the mantle is via emplacement of granitic intrusions which occurs in roughly equal weight proportion. From https://en.wikipedia.org/wiki/Granite

Quote

Chemical composition[edit]

A worldwide average of the chemical composition of granite, by weight percent, based on 2485 analyses:^[14]

SiO2	72.04% (silica)
Al2O3	14.42% (alumina)
K2O	4.12%
Na2O	3.69%
CaO	1.82%
FeO	1.68%
Fe2O3	1.22%
MgO	0.71%
TiO2	0.30%
P2O5	0.12%
MnO	0.05%

Chemical weathering of rock minerals follow the Goldich Dissolution Series 

Quote

Discontinuous Series

Continuous Series

High

Olivine

Plagioclase (Calcium rich)

Pyroxene

Amphibole

Biotite (Black Mica)

Plagioclase (Sodium rich)

Relative Weathering potential

Orthoclase

Muscovite (White Mica)

Quartz

Low

This indicates that the potassium rich minerals (orthoclase, muscovite, biotite) are relatively resistant in comparison to sodium rich minerals (sodic plagioclase, some amphiboles and clinopyroxenes). By logical extension, sodium is over-represented in sea water and evaporites, whereas potassium is over-represented in detrital sands and sandstones. 

[swansont]

3 hours ago, exchemist said:

Aha, that makes a difference, certainly!

But there is a factor of 30 to account for.

Yes, that only accounts for a little bit of the difference. It’s likely all the factors mentioned contribute. Several factors of 2 or 3 rather than one big one.

[exchemist]

10 hours ago, sethoflagos said:

If this helps, the primary route for transport of sodium and potassium into the earth's crust from the mantle is via emplacement of granitic intrusions which occurs in roughly equal weight proportion. From https://en.wikipedia.org/wiki/Granite

Chemical weathering of rock minerals follow the Goldich Dissolution Series 

This indicates that the potassium rich minerals (orthoclase, muscovite, biotite) are relatively resistant in comparison to sodium rich minerals (sodic plagioclase, some amphiboles and clinopyroxenes). By logical extension, sodium is over-represented in sea water and evaporites, whereas potassium is over-represented in detrital sands and sandstones. 

This is excellent stuff, which I feel sure must be on the right track.

The next question, in my rusty chemical mind, is why this should be so. I suspect something to do with the size of the "cages"  in the crystal lattice formed by silica tetrahedra, in complex silicate minerals. I feel sure the difference must be to do with the difference in size of the cations in some way. Generally speaking large cations are more stable in structures with a counterion network that has larger interstices for them to occupy. It may be that Na+ "rattles around" in these structures, has less stability and can slowly diffuse out as they weather.  As I recall, these minerals have silica tetrahedra that can be joined either at vertices or along edges, giving different sized holes. If I have time later I'll see whether I can find anything further along these lines to explain cation preferences in these minerals.

Edited January 20 by exchemist

[sethoflagos]

13 hours ago, exchemist said:

Aha, that makes a difference, certainly!

But there is a factor of 30 to account for.

@swansont's reference is dominated by the composition of the earth's mantle. 

Quote

 

Composition of the Earth's upper mantle (depleted MORB)^[16][17]

Compound	Mass percent
SiO2	44.71
MgO	38.73
FeO	8.18
Al2O3	3.98
CaO	3.17
Cr2O3	0.57
NiO	0.24
MnO	0.13
Na2O	0.13
TiO2	0.13
P2O5	0.019
K2O	0.006

 

This is indeed the ultimate source of sodium and potassium at the earth's surface, and the sodium:potassium ratio is strikingly similar to the marine ratio. However, I think this is coincidental. Except for odd cases such as a notable ocean bed exposure off the Cape Verde islands, surface waters and upper mantle rocks are not significantly in contact. 

Rather, there are a number of fractionation processes in the vicinity of the crust-mantle boundary that enrich the alkali metal content and deplete the mafic materials (Mg, Fe predominantly), starting with emplacement of basalt/gabbro at constructive plate margins and culminating with the granitisation of the base of the continental crust at destructive plate margins. 

Even this is quite a simplification as there are a number of other processes involved most of which are not fully understood, so I tend to take a first order engineering approximation of the mantle injecting granite into the continental crust. It's easier on my head and the overall mass balance still works.

The concentration factors of 30 for sodium and 600-ish for potassium reflect their relative preferences for a granitic environment over a peridotite environment.

[exchemist]

8 minutes ago, sethoflagos said:

@swansont's reference is dominated by the composition of the earth's mantle. 

This is indeed the ultimate source of sodium and potassium at the earth's surface, and the sodium:potassium ratio is strikingly similar to the marine ratio. However, I think this is coincidental. Except for odd cases such as a notable ocean bed exposure off the Cape Verde islands, surface waters and upper mantle rocks are not significantly in contact. 

Rather, there are a number of fractionation processes in the vicinity of the crust-mantle boundary that enrich the alkali metal content and deplete the mafic materials (Mg, Fe predominantly), starting with emplacement of basalt/gabbro at constructive plate margins and culminating with the granitisation of the base of the continental crust at destructive plate margins. 

Even this is quite a simplification as there are a number of other processes involved most of which are not fully understood, so I tend to take a first order engineering approximation of the mantle injecting granite into the continental crust. It's easier on my head and the overall mass balance still works.

The concentration factors of 30 for sodium and 600-ish for potassium reflect their relative preferences for a granitic environment over a peridotite environment.

I'm not sure I follow this. Surely both Na and K prefer granite to peridotite, the latter being ultramafic (Mg/Fe), don't they?  

[sethoflagos]

34 minutes ago, exchemist said:

I'm not sure I follow this. Surely both Na and K prefer granite to peridotite, the latter being ultramafic (Mg/Fe), don't they?  

Sodium is relatively happy in some mafic minerals, particularly clinopyroxenes. I suspect potassium simply won't fit in that lattice.

In a feldspar environment, the alkali metal hole is clearly big enough for potassium and maybe this gives a little more stability to the potassium version under most surface conditions. However relative weathering rates show a high pH and temperature sensitivity so perhaps there are multiple mechanisms at play.

[exchemist]

On 1/20/2024 at 12:41 PM, sethoflagos said:

Sodium is relatively happy in some mafic minerals, particularly clinopyroxenes. I suspect potassium simply won't fit in that lattice.

In a feldspar environment, the alkali metal hole is clearly big enough for potassium and maybe this gives a little more stability to the potassium version under most surface conditions. However relative weathering rates show a high pH and temperature sensitivity so perhaps there are multiple mechanisms at play.

OK, however I am  not clear now what the relative proportions of Na and K would have been in the Earth's crust , or crust + mantle, before the oceans formed. Without that information it seems hard to determine whether the reason for the difference in concentration of today's seawater is due mainly to the original composition of the rocks or to the differential leaching that we have been discussing. 

By the way, the poster who asked the original question does not seem to have returned. Perhaps he is watching and chuckling to himself as we struggle with it.😁

 

[sethoflagos]

6 hours ago, exchemist said:

OK, however I am  not clear now what the relative proportions of Na and K would have been in the Earth's crust , or crust + mantle, before the oceans formed. Without that information it seems hard to determine whether the reason for the difference in concentration of today's seawater is due mainly to the original composition of the rocks or to the differential leaching that we have been discussing.

On reflection, I'm inclined to think that it's a bit of each. The salinity of the Hadean oceans would no doubt have reflected the mantle Na:K ratio which in turn is reflected in the Na:K estimated ratio of the solar system. So perhaps the similarity in current times isn't as much of a coincidence as I suggested in an earlier post.

However, with the onset of plate techtonics in the mid-Archaean, the two following quotes, taken together indicate that both surface waters and their salt inventory are being continually recycled to and from the mantle:

On 1/20/2024 at 12:27 AM, TheVat said:

In fact, the oceans would be even more saline were it not for subduction.

 

Quote

Experiments have documented that olivine at high pressures (12 GPa, the pressure at depths of about 360 km (220 mi)) can contain at least as much as about 8900 parts per million (weight) of water, and that such water content drastically reduces the resistance of olivine to solid flow. Moreover, because olivine is so abundant, more water may be dissolved in olivine of the mantle than is contained in Earth's oceans.^[17]

 

I guess the turnover time is at least oto 1 billion years so it's a very slow process and almost certainly not at equilibrium. But it is clear that there is some continuous limited exposure to mantle cation ratios that would act as a negative feedback loop tending to restore primitive values.

Which leaves us with having to deal with why continental crust is so markedly different with a near unity Na:K mass ratio.

On 1/20/2024 at 12:51 PM, exchemist said:

Surely both Na and K prefer granite to peridotite, the latter being ultramafic (Mg/Fe), don't they?

Arguably so, but at constructive plate margins, there is no neighbouring granite for them to migrate to. Rather we have a fractionation process that might be roughly summarised as:

                                   3 Peridotite => 1 Dunite + 2 (Gabbro + Basalt)

Dunite is over 90% olivine, offers very limited hospitality to Na (maybe in some residual pyroxene) but as far as I can tell none to K. So most of the Na and all of the K plus the rest of the more volatile components creates a gabbro melt that ascends to fill the gap between the separating plates.

At this point I wanted to give some indicative Na:K ratios for oceanic crust gabbro/basalt. Only to find lots of field data showing Na:K ratios up to 50+ for gabbro and 2 for basalt. The common understanding is that basalt is simply rapidly cooled gabbro. Clearly this is somewhat of a simplification. On the face of it, the fractionation process is not the unitary division I've just described, but some multistage fractionation process that preferentially shunts Na into the gabbro levels and K into the basalts. I'd be interested in your views on this.

Anyway, it appears that we have the K concentrated in the upper basaltic levels of the oceanic crust, possibly in the form of a feldspathoid such as leucite, the Na evenly spread between the basalt and lower gabbro horizons within clinopyroxenes such as augite, and below the crust mantle boundary a zone of depleted cumulate dunite.  

At the other end of the conveyor hopefully the picture is a little clearer. As the old oceanic plate is subducted, it eventually reaches a region of high temperature, high pressure which (give or take a little controversy) is where it is converted to a highly metamorphosed rock type called eclogite, mainly comprising pyralspite garnet and a sodium rich clinopyroxene called omphacite. 

With the K being concentrated in the contact zone of the upper levels of the oceanic crust, and an incompatible mineral assemblage forming beneath it, the path for transport of the vast majority of it into the continental crust appears non-problematic. Na on the other hand has the option of joining with the K to progress upwards, or staying in situ within a compatible mineral assemblage. The nett result seems to be a roughly equal mass diffusion of sodium and potassium into the lower levels of continental crust.

 I basically have zilch documentary evidence to support the overall picture of this mass balance, but it makes some sort of sense to me and my researches haven't yet thrown up anything that cotradicts it significantly. Hope it helps.

6 hours ago, exchemist said:

By the way, the poster who asked the original question does not seem to have returned. Perhaps he is watching and chuckling to himself as we struggle with it.😁

I'm not going to lose sleep over that 🙂 

Edited January 21 by sethoflagos
typo

[WillyEngland]

Ok, belated, thank you very much for your answers!

Did you come to a conclusion? 🙂

[sethoflagos]

1 hour ago, WillyEngland said:

Did you come to a conclusion? 

The Na:K mass ratio order of magnitude appears to be established by the ~11:1 ratio in the Earth's mantle. This probably manifests predominantly as hydrothermal vents entering the oceans at mid-oceanic ridges.

Beyond that it gets complicated as the elements take different routes in their respective cycles. 

However, a huge amount of ancient marine potassium has become preferentially locked up in the granites of the Earth's continental crust which must account for some significant depletion of the residual potassium in the ocean relative to sodium. It isn't the complete picture but it may be a good part of it.