Kevin Practicals Flashcards Preview

Petrology > Kevin Practicals > Flashcards

Flashcards in Kevin Practicals Deck (33)
Loading flashcards...

182Hf-182W chronometry of core formation

Bulk planet = chondrites = 1 Hf/W
Core = low Hf/w
Mantle = high Hf/W

Simplest interpretation is that the Earths core has a mean age about 30 Myr after the start of the solar system


Outline the principal limitations of using the Hf-W isotope system

Earth grew by accretion of smaller planets and planetesimals, so the W isotope composition of the silicate Earth is dependent upon the (1) history and nature of accretion (that is, how long accretion occurred and the composition of planetesimals accreted to the Earth) (2) the degree of metal-silicate equilibration with Earth’s mantle-core.


In the abstract of their paper Kruijer et. (2015) argue that tungsten isotopes are useful for testing the ‘late veneer’ hypothesis,
(i) Outline the two reasons that they give in the abstract.

If the ‘late veneer’ did occur,
(ii) How would this have affected the 182W isotope composition of Earths mantle.
(iii) what 182W would you expect the Moon to possess relative to the Earth

(i) First, the late veneer material would have had a different W isotope composition to the bulk silicate Earth. Second, proportionally more material was added to the Earth than to the Moon.

(ii) addition of chondritic W with a relatively unradiogenic 182W composition would lower the isotope composition of the silicate mantle to less radiogenic values.

(iii) If the Moon received proportionally less late veneer than Earth then it should possess a slightly more radogenic 182W isotope composition.


(i) What is the problem with 182W isotope data for the Moon that existed prior to the study of Kruijer et al. (2015) and (ii) why is measuring the lunar 182W isotope composition more complicated than for rocks on Earth?

(i) Data for lunar rocks prior to the study of Kruijer et al (2015) is not precise enough to identify a signature for the late veneer, in particular, not precise enough to distinguish a difference between the Earth and the Moon.

(ii) Measuring the 182W isotope composition of lunar rocks is complicated by cosmic-ray induced secondary neutron capture reactions. These include 182W production via neutron capture by 181Ta, and neutron capture induced burnout of 182W, which lowers 182W.


How do the authors use (i) KREEP-rich basalts, and (ii) 182W vs 180Hf data to determine which lunar rocks give a robust 182W value?

(i) KREEP-rich basalts all have the same Ta/W ratio, so any variations in 182W due to neutron capture by 181Ta, are due to differences in exposure to cosmic rays, rather than differences in the Ta/W ratio.

(ii) 180Hf is only produced by cosmic-ray induced neutron capture, so any positive covaraition of 182W with 180Hf must be due to the same process (Figure 1 in the paper). And samples with an 180Hf = 0 must be little affected by neutron capture


What is the 182W value determined for lunar rocks, and how does it differ from that of the bulk silicate Earth.

The 182W value of the Moon is +0.27±0.04, and this is the difference between the Moon and the silicate Earth (Figure 2 in the paper)


Is the difference between the 182W value of the Earth and Moon consistent with the addition of a ‘late veneer’

The difference between the Earth and the Moon is consistent with the addition of 182W- depleted material (chondritic) to the Earth, with a total mass consistent with that derived from the HSE abundances in Earths mantle (Figure 3 in the paper). (so, prior to addition of the late veneer the Earth and Moon had an indistigusihable 182W isotope composition).


The recent study of Thiemens et al. (2019) suggests that the Earth and Moon have different Hf/W ratios and therefore that the difference in 182W, simply indicates that the Moon was formed before 182Hf became extinct. How do they explain the difference in Hf/W between the Earth and the Moon?

Either (i) The Moon-forming event occurred while Earth’s core was still forming and 182Hf is still decaying. Where increasingly oxidizing conditions later lower the BSE Hf/W, because W becomes less siderophile, or (ii) the formation of a small lunar core scavenged W from the BSM, which then increased its Hf/W.


(a) What is the unusual chemical characteristic of komatiites ?
(b) What does such chemical composition tell us about melting degree and how is this degree of melting usually explained ?
(c) How might komatiites provide reliable information on the chemistry of the deep mantle?

(a) High MgO content (and textures indicative of rapid cooling)

(b) High melting degrees, up to 40% (is this right?). Usually explained as the result of melting at high temperatures (or high water content)
- Earth was hotter so intercepted the solidus earlier

(c) High degrees of melting so reliable information on mantle compositions, and sample large volumes of the upper mantle
- shows mixing trend in the mantle


Using Figure 2a from the study of Maier et al (2009) (shown below) explain what they discovered that had not been observed in previous work.
Pt (at 25% MgO) versus age of komatiites

This study shows that younger komatiites are progressively more enriched in Pt at a given MgO content, and that the most signficiant HSE depletion is found in the oldest rocks, and concentrations similar to present-day were attained by 2.9 Gyr.
Addition of platinum maybe from late veneer and then mixing later = systematic increase


Outline the model favoured by the authors to explain the HSE depletion of early Archean komatiites ?

The preferred model assumes that the Earth’s mantle was initially depleted of HSEs by core formation, within the first 100 Myrs after its formation, followed by progressive re-enrichment with HSE in response to late addition of meteoritic material.

If komatiites originate from melting in the deep mantle then this signal reflects that of the deep melting, indicating that the progressive increase is due to mixing-in of the late veneer into the upper mantle and transport to the deeper mantle.


Highly siderophile elements and tungsten isotopes in 3.9 billion year old rocks

The study of Dale et al. (2017) presents highly siderophile element and tungsten isotope data for 3.9 billion year old volcanic rocks from Isua, Greenland.

(a) What do they observe for the measured tungsten isotope ratio 182W compared to that of modern terrestrial samples (their Figure 4, shown below) (b) When in Earth’s history must the isotope composition of source of the Isua volcanics have been generated ?

(a) The measured 182W/184W for the Isua volcanics is about 13 parts per million higher than the modern terrestrial samples.

(b) Because the half-life of 182Hf is about 9 Myr then this difference must have been generated in the first 50 Myr of Earth’s history. Which implies that the part of the mantle which melted to produce these volcanics must have survived 500-600 Myr. That is, from the first 50 Myr of Earth’s formation to 3.9 Million years (when the volcanic rocks were erupted at the Earth’s surface_


(a) Explain their preferred interpretation of the difference in 182W/184W between the Isua volcanics and the modern mantle.

(b) Explain whether this model fits with the HSE data from Maier et al. (2009) ?

(a) In their model Dale et al. (2017) consider that the difference relates to the late veneer. The mantle source sampled by the Isua volcanics represents a part of the mantle that has been relatively unaffected by the late veneer, whereas addition of the late veneer (with a chondritic 182W/184W isotope composition) has lowered the value of the modern mantle by 13 ppm.

(b) Tungsten is a moderately siderophile element and so will only be affected to a very small extent by addition of the late veneer. In contrast, HSEs are highly siderophile and should be srongly depleted by core formation, and subsequently enriched by the late veneer. The data of Maier et al. (2009) suggests that the earliest mantle was depleted in HSEs, consistent with the model proposed by Dale et al. (2017).


The figure below (their Figure 6) shows the e182W data and HSE abundance data for Isua volcanics and the Moon.

Are these results consistent with the interpretations of Kruijer et al. (2015) from last weeks practical ?

Broadly speaking yes, the value e182W value for the Moon could be taken to represent the composition of the mantle prior to the addition of the late veneer. The mantle source of the Isua volcanics is a little heavier, reflecting the addition of some proportion of the late veneer. While the modern mantle has the full complement of HSEs.


OIB plumes and H

Ocean island basalts (OIB) derived from mantle ‘plumes’ sometimes have 3He/4He ratios that are much higher than mid-ocean ridge basalts (MORB). High 3He/4He and solar-like Ne isotope ratios reflect a higher proportion of primordial volatiles in the source region, and indicate that ‘plumes’ tap a deep mantle reservoir that is significantly less degassed than the asthenospheric mantle.


The study of Rizo et al. (2016) presents new 182W and 142Nd isotope data, and highly siderophile element (HSE) abundance data, for the picritic flood basalts from Baffin island, and the Ontong Java plateau.

Q. 1. (i) What does the 182W isotope data for these flood basalts show?
(ii) what does this imply for the mantle source of these flood basalts?
(iii) How do these 182W values compare with those for the Moon ?
(Note that the data in this study are given as ppm, where 182W = 182W x 10-2)

(i) Variable 182W values from +10 to +50 ppm, all higher than modern mantle derived lavas.
- A positive anomaly compared to normal terrestrial material

(ii) Basalts sample parts of the mantle that must have been formed in the first 60 Myr of earth’s history, and that these have been isolated from the rest of the mantle (not mixed in)

(iii) Some less radiogenic than the Moon (182W = +28 ppm, Kruijer et al. 2015) but some higher.


(i) What does the HSE data (shown in Figure 2) for the same samples show and,
(ii) how does the data compare with that for modern lavas (same Figure),
(iii) Does the HSE data supprt the idea that these flood basalts were derived from a mantle reservoir that was isolated from the ‘late veneer’ (like the Isua volcanics shown in Figure 2 in last weeks practical)

(i) The HSE data are similar 1-2 orders of magnitude less than chondritic (like average mantle)

(ii) Similar to other modern lavas, such as the mantle picrites shown in Figure 2

(ii) No because the mantle source appears to have the full complement of HSEs from the ‘late veneer’. Also 182W = +50 is even more radiogenic than the Moon


What other processes might cause the fractionation of Hf/W?

1. Core formation – W is siderophile, Hf is lithophile
2. Crystallisation of the magma ocean/mantle melting – W is more incompatible than Hf


Why are silicate fractionation processes (crystallisation/partial melting) as a means of causing Hf/W fractionation, ruled out by the 142Nd data?

Because magma ocean crystallisation or partial melting of the mantle would fractionate Sm/Nd and result in the generation of anomalous 142Nd values.


Summarise the preferred model for the creation of mantle reservoirs with distinct W isotope compositions on Earth (outlined in Figure 3 of Rizo et al. (2016).

1) Formation of the core – leaving residual mantle with high Hf/W and radiogenic 182W (182W = +50 or higher).

2) Later impact (perhaps Moon forming impact) creating a mantle domain with lower 182W value

3) Mantle reservoirs with distinct 182W values – some mixing but deep mantle preserves the most primitive signature.

4) Late veneer adds HSE’s and some W, decreasing 182W of all reservoirs.

Problem: How does earliest reservoir acquire high HSEs, i.e. indistinguishable from modern mantle, and why the same – unlikely to be a coincidence


Describe the two mechanisms that have been proposed to account for the elevated 142Nd in the Earth. That is, those models that assume that the elevated 142Nd is only due to the decay of 146Sm.

With each model highlight how each specifically leads to an elevated 142Nd, and outline some of the potential problems.

Two general mechanisms: (1) crustal erosion during collisional accretion or (2) formation of an enriched hdden reservoir on Earth.

(1) Melting of the mantle results in a crust which is enriched in incompatible elements, including Sm and Nd, and has a low Sm/Nd ratio. The complementary mantle has a high Sm/Nd. During accretion, mass can be lost as well as gained and this process may preferentially remove the outer layers of a growing planet. As a consequence, the final planet ends up with a super-chondritic Sm/Nd, having disproportionately lost low Sm/Nd crust. Since accretion occurs with the first few 10s millions of years of the solar system, compared to the half-life of 146Sm, the high Sm/Nd ratio of silicate material left on Earth will evolve to an elevated 142Nd composition.

Problems: (i) loss of incompatible heat-producing elements (Th, U and K), so heat-production is diminished, and primordial heating must have been even greater to account for present-day heat flux (ii) Need to fractionate Sm/Nd during formation of basaltic crust, not seen at present day in MORB, or on other early planetesimal crusts, such as Vesta.

(2) The hidden reservoir requires an early formed enriched reservoir to be formed from the convecting mantle. In detail, it has been argued that this could be (i) Early formed crust that is enriced (with low Sm/Nd) but foundered and sunk to the base of the mantle. (ii) Enriched melt, left from the crystallisation of the magma ocean, trapped at the base of the mantle (with low Sm/Nd). (iii) Partitioning of incompatible elements into the metallic core, experimental data suggests that both Sm and Nd can be partitioned into sulfide or S-rich metal (with low Sm/Nd).

Problems: (i) No evidence thermal or chemical evidence for an enriched reservoir in the deep mantle (ii) models suggest that convection in the liquid outer core could not occur with a hot layer in the deep mantle. (iii) Deep layer would have to have survived the Moon-formaing impact (iv) incompatible elements may go into core, provides a heat-supply for convection, but difficult to prove or disprove.


Problems with hidden reservoir

How can you hide it? When 30-40% of heat producing elements yet no temp signature
Problems of geodynamo
Geochemical issue - no evidence of it


Burkhardt et al. (2016).

(1) Radiogenic decay: 146Sm --> 142Nd + 3He (radiogenic decay)

Or the result of nucleosynthetic variability, for example;

(2) s-process neutron addition: 141Pr + neutron --> 142Nd + beta particle

Q. 2 Decsribe how the authors propose that they can distinguish between these two process and, with reference to Figures 1 and 3 in their paper, outline their preferred interpretation.

Radiogenic decay only influences 142Nd, if the Earth where super-chondritic in 142Nd as a result of decay, then only 142Nd in the Earth should differ from the chondrites.

Whereas, if the 142Nd variability is due to nucelosynthetic processes then there should be a variability for all Nd (and Sm) isotopes between Earth and chondrites.

Figure 1 hints at differences for 142Nd, 145Nd and 150Nd, Figure 3 shows that the variations in ordinary and carbonaceous chondrites are consistent with a deficit in s-process nuclides relative to the Earth and enstatite chondrites.


Burkhardt et al. (2016) claim that these new data ‘rule out’ or ‘refute’ those process inoked to explain the 142Nd variations by radiogenic decay. Discuss conditions under which the mechanisms discussed in querstion 1 are not ruled out.

Does not rule out either process - just means they are less likely to be the reason for differences in Sm/Nd

(1) collisional erosion – crust enriched in incompatible elements, but no fractionation in Sm/Nd

(2) Hidden resrvoir – likewise, no fractionation in Sm/Nd


The authors document a variability in non-radiogenic Nd isotope compositions between different meteorite groups. Explain the signficance of these isotope variations for the structure of the solar nebula and the accretion of the Earth (outlined in the last paragraph of the paper).

The authors argue that chondrites are the closest proxy for the elemental composition of the Earth. But that they cannot be the actual building blocks of the Earth because they are deficient in a pre-solar component that contains s-process material.

The s-process deficit increases from enstatite, to ordinary, to carbonaceous chondrites, indicating that the distribution of presolar matter in the protoplanetary disk varied either as a function of heliocentric distance or over time.

On the basis of the higher volatile content of carbonaceous chondrites it is often argued that these meteorites are derived from bodies outside the snow line. In contrast the highly reduced esntatite chondrites are more likely to have formed in the hot inner portion of the disk. The authors suggest that this isotope heterogeneity might reflect differences in thermal processing of stellar derived dust or that distinct compositions of material were added to the disk at different times.


(a) What does the correlation between Rb/Sr and SiO2 in crustal rocks indicate? (b) How can this relationship be used to infer the origin of crustal rocks ?

- increase in Rb/Sr with increasing SiO2 in crustal rocks
- Can use the Rb/Sr ratio to infer the nature of the crust.


The study of Bruno Dhuime, Andreas Wuestefeld & Chris J. Hawkesworth (Nature Geosciences 8, 552–555 (2015)) suggests that Rb/Sr ratios can be used as a proxy for both the silica content and the thickness of the continental crust, and this information provides some insight into the eviolution of the continental crust.
What assumptions have the authors made in this simple two-stage model ?

- that crustal remelting only occurs once, whereas may occur multiple times
- that Rb/Sr have remained a closed system during stage 1 & 2, no Rb or Sr loss or gain
- that Sm/Nd are not fractionated during the remelting event


(a) How does the Rb/Sr ratio of juvenile crust change with time? (b) What do the authors deduce from this change in Rb/Sr ratio?

- Low Rb/Sr ratio until 3 Ga, then a systematic increase until ca. 1 Gyr, small decrease since then
- Prior to 3 Ga crust dominantly mafic, since that time systematic increase in Rb/Sr and ny implication SiO2 content indicates that crust is increasingly differentiated


How do the authors explain the link between Rb/Sr and SiO2 with crustal thickness shown in Fig. 3 (above)

- increased differentiation in areas of thicker continental crust


What happens to the relationship between crustal thickness and Rb/Sr and SiO2 at crustal thicknesses greater than ~55 km? How can these be explained?

- Rb/Sr much higher in crust >60 km thick, while SiO2 is restricted to ~60 wt. %. In situ melting of the (lower) crust at greater crustal thicknesses increases proportion of more incompatible elements in melts (such as Rb), therefore increases the Rb/Sr ratio. Melting of lower crust (thought to be less felsic than upper crust, made of residual cumulates) which is less SiO2 rich, therefore restricts SiO2 increase with increasing crustal thickness.