Low thermal conductivity of iron-silicon alloys at Earth’s core conditions with implications for the geodynamo

Earth’s core is composed of iron (Fe) alloyed with light elements, e.g., silicon (Si). Its thermal conductivity critically affects Earth’s thermal structure, evolution, and dynamics, as it controls the magnitude of thermal and compositional sources required to sustain a geodynamo over Earth’s history. Here we directly measured thermal conductivities of solid Fe and Fe–Si alloys up to 144 GPa and 3300 K. 15 at% Si alloyed in Fe substantially reduces its conductivity by about 2 folds at 132 GPa and 3000 K. An outer core with 15 at% Si would have a conductivity of about 20 W m−1 K−1, lower than pure Fe at similar pressure–temperature conditions. This suggests a lower minimum heat flow, around 3 TW, across the core–mantle boundary than previously expected, and thus less thermal energy needed to operate the geodynamo. Our results provide key constraints on inner core age that could be older than two billion-years.

The results are provocative because the author advocate for a much lower thermal conductivity than that obtained from theory. This reduction in conductivity has enormous consequences for the thermal evolution of the core and the origin of the Earth's magnetic field. The authors have done a good job conveying the significance of their experiments and providing insightful and quantitative estimates for the implications.
The most important question is whether the experimental measurements and analysis are sufficient to reliably support the authors' conclusions. The authors describe their efforts to propagate uncertainty through the analysis of their measurements. This is important because the thermal conductivity must be extracted from a modelling procedure. Uncertainties in the model parameters will affect the outcome. The authors suggest that the uncertainty is about 10-20%. I wonder if these error estimates are too low. Figure S4 shows the temperature dependence at various pressures. I would be surprised is the erratic pressure dependence is real, which suggests that these variations are reflection of experimental uncertainty. Still, the results are sufficient to argue for lower thermal conductivity, even with larger experimental uncertainties.
There is little doubt that this study will motivate future work, either to reproduce these results or to extend the experiments to other light elements. It seems likely that silicon will be present at some level in the core, so the results should be relevant moving forward. I would encourage the authors to make sure that their error estimates will stand the test of time.
The present study could be published in its current form, although I would encourage the authors to be clear early in the paper that the experiments are carried out on solid samples. This detail emerges once the experiments are described, but it is sufficiently important to mention in the abstract. Allowing for the likely changes due to melting only reinforces the authors' conclusions.
Reviewer #2 (Remarks to the Author): This paper report the results of the direct determination of thermal conductivity of FeSi alloy at high pressure, and revealed that it is incorporation of Si in hicp-Fe significantly reduces the thermal conductivity of 20 Wm-1K-1. Which implies the heat flux at the CMB reduces to around 3 TW.
There is a discrepancy of a high thermal conductivity inferred from the electrical conductivity, whereas a low thermal conductivity from the transient heating laser method. These discrepancies are very significant and provide the different views for the formation of the inner core and the early magnetic field.
The senior author is an expert of the thermal conductivity measurements at room temperature by using TDTR technique and the results are considered to be reliable. The thermal conductivity at high pressure and high temperature was made by the TH method, which is also considered to be reliable since the co-authors have significant experience of the measurements. This is a clear and concise paper showing lower thermal conductivity of the Fe-Si alloys. Discussion is clear and arguments and conclusion is reasonable. Implications for the formation of the inner core and the core dynamo is very important. Therefore this paper can be suitable for this journal. However, there are so many incomplete figures and captions. Therefore, the authors should rewrite the figures and the captions completely. We cannot accept it in the present version. Major revision is needed for publication.
Specific comments especially on the figures.
Page 2, First paragraph and Page 6 the first paragraph of discussion: I would like to make some discussions on the mechanism of heat transfer at very high pressure and temperature. Generally heat is transferred by three mechanisms, such as convection heat transfer, thermal conduction which is measured here experimentally, and radiation. I am wondering that the radiation heat transfer may not be negligible and this mechanism might be important and can increase the thermal conductivity significantly in spite of the low thermal conductivity determined here. We need some discussions on the heat transfer mechanism in the core.
Page 5, 2nd paragraph: The authors used two techniques, TDTR at 300 K and TH at high temperature. However, there is no discussion on the consistency of these technique. I would like to see some discussions on the consistency of the two data-sets. The data on TDTR at room temperature might extrapolate theoretically, and could compare the data by TH. The Si content dependency for TDTR and TH are consistent? Figure S3: There is no explanation of the blue curve. No explanation for FE (finite-elements) calculation. FE calculations should be FE calculation. Please fix this figure. Figure S4: The change of the FeSi thermal conductivity shows strange pressure dependency at high temperature, i.e. there is a maximum at 106 GPa. Provide a reasonable explanation of the pressure dependency at high temperature of 2500-3000 K. Is it due to a two phase mixture of BCC and HCP? If the anomalous change is due to uncertainty of the TH method, please provide uncertainty for the plot for the measurements. The top side and bottom side of the sample is related to the pulsed side and opposite side given in Figure S3? Figure S6: Provide a reference for the phase diagram of FeSi alloy. It may be ref. 12, but it is confusing that black triangles and open circles are the literature data for phase identification by ref. 12, but NOT the data points for previous thermal or electrical conductivity measurements. The measurement made with TH experiments was for the mixture of hcp and bcc, and not a single phase. Please explain the differences of your data points shown in yellow stars and brown stars. Do you show the two series of measurements? What is "Bottom temperature" in this figure?
The solid lines for the boundaries of bcc, hcp, fcc may be those of pure Fe instead of Fe0.85Si0.15. Therefore it should be deleted.
Reviewer #3 (Remarks to the Author): Review on "Low thermal conductivity of iron-silicon alloys at Earth's core conditions: implications for the geodynamo" by Hsieh et al.
In this study, the authors measured thermal conductivity of Fe and Fe-Si alloys at high pressures and room temperature and at high pressures and high temperatures by means of the transient heating method in DAC. They report that the thermal conductivity of Fe-Si alloy is lower than pure Fe at high temperatures and ambient temperature. They also report that thermal conductivity of 20-60 W/mK at high pressures and temperatures, which is much lower than previously estimated using electrical conductivity based on the Wiedemann-Franz's law. According to these results, they concluded that the inner core is older and that the initial core-mantle-boundary is colder than previously estimated.
I have several scientific comments on this manuscript.
1. I think that their main conclusion that the Fe-Si alloys have low thermal conductivity than pure Fe is invalid at high pressures and temperatures. Figure 2 demonstrates that the thermal conductivities of pure Fe and Fe-Si alloys are simply indistinguishable at high temperatures, although it may be the case at ambient temperature as shown in Fig. 1. Table S1-S3, the measurements for Fe0.96Si0.04 and Fe0.93Si0.07 were conducted at a temperature of 2050 K, whereas those for Fe0.85si0.15 were at temperatures of 2300 to 3100 K. Therefore, this study does not present consistent experimental results to show the compositional dependence of thermal conductivity of Fe-Si alloys. 3. The thermal conductivity of Fe and Fe-Si alloys increases by adding percent orders of Si to Fe, whereas it decreases by adding 15at% of Si. This would indicate that the effect of Si addition is monotonic, and therefore, they cannot argue core thermal conductivity simply using that of Fe0.85Si0.15 alloy. 4. Figure S4 shows that the thermal conductivity of Fe0.85Fe0.15 increases with increasing temperature at pressures of 47~106 GPa, whereas it is independent from temperature at higher pressures. The author need to argue the mechanism to cause these phenomena. 5. The outer core temperatures are more than 4000 K. The present temperature range is insufficient for arguing the core conductivity, because the temperature effect on thermal conductivity does not seem simple as argued above. 6. I doubt that argument of the core thermal history based on thermal conductivity of core materials is essential. The core has cooled due to heat release across the CMB. If there were no heat production in the mantle, the released heat should be equal to the total surface heat flow. Thus, the core coolingrate should be controlled by the surface heat flow of the Earth not by the heat transfer in the core. This argument is strengthened by the fact that the thermal conductivity of rocks is much lower than that of Fe alloys. Therefore, it is the thermal-boundary-layer structure of the bottom of the Earth that controls the amounts of heat released from the core to the mantle. The amounts of heat transfer within the core should be adjusted by verbosity of thermal convection in the outer core. I know numerous studies argued the cooling history of core based on Fe and Fe-alloy thermal conductivity as shown in Table 1 and 2, but I think that they are useless papers.

As shown in
I have several comments on the presentation. 7. Although the abstract states "15 at% Si alloyed in Fe substantially reduces the conductivity by about 2 folds at 132 GPa and 3000 K.", there is no such argument in the main text. 8. The authors used single-crystal and polycrystalline samples. They used both samples for measurements at ambient temperature. However, it is not shown in Fig 1 which sample was used. Furthermore, it is unclear whether single-crystal samples were used for high-temperature measurement. They also should show which sample was used in each measurement in Table S1-S3. 9. Page 10. I do not understand why there is a description regarding MORB. 10. Page 14. It is clear that high-temperature measurements should be much more important than ambient-pressure measurements in view of geophysics. Therefore, the descriptions regarding hightemperature measurement should have larger amounts than those regarding ambient-pressure measurements. I wonder what fitting parameters were used for data analyses of high-temperature measurements. The authors should provide a similar table to Table S4 for the high-temperature measurements. 11. Page 28. I am afraid that Figure S3 is not used anywhere in the text. 12. Page 32. The structures of these tables are confusing. One row has six columns and each of three columns describes one measurement. Even though space-consuming, each row should describe only one measurement.
As seen from the comments above, I am not supportive to this article. The paper lacks convincing experimental data to argue the thermal conductivity of core materials at high pressures and temperatures. More essentially, thermal conductivity of I also point out that this paper is organized unreasonably. I strongly suggest the authors to reconsider their measurement and arguments, and to reorganize the whole parts of the article.
We thank you and the three reviewers for helpful comments. In what follows, all reviewers' comments are in italics and our point-to-point responses are in normal fonts with blue color. Changes in our revised manuscript are also labeled by blue color. We have added a significant amount of materials and information, which, we believe, have fully addressed the reviewers' concerns and questions.
Report of Referee # 1 1. The most important question is whether the experimental measurements and analysis are sufficient to reliably support the authors' conclusions. The authors describe their efforts to propagate uncertainty through the analysis of their measurements. This is important because the thermal conductivity must be extracted from a modelling procedure. Uncertainties in the model parameters will affect the outcome. The authors suggest that the uncertainty is about 10-20%. I wonder if these error estimates are too low. Figure S4 shows the temperature dependence at various pressures. I would be surprised is the erratic pressure dependence is real, which suggests that these variations are reflection of experimental uncertainty. Still, the results are sufficient to argue for lower thermal conductivity, even with larger experimental uncertainties. 2. There is little doubt that this study will motivate future work, either to reproduce these results or to extend the experiments to other light elements. It seems likely that silicon will be present at some level in the core, so the results should be relevant moving forward. I would encourage the authors to make sure that their error estimates will stand the test of time.
As we respond to the comment #1 above, we have confirmed our experimental error estimates and clarified the discussions of the experimental uncertainties in the revised manuscript where appropriate. On the other hand, the exact temperature dependence of the thermal conductivity of Fe-Si alloys at high pressures with different Si content remains relatively uncertain.

The present study could be published in its current
As we reported in Fig. 2   To discuss the consistency of the TDTR and TH results, we have added a new paragraph (as we stated above) in line 148-165 on page 6-7 of the section "Discussion".
In addition, to add further comment on the effect of temperature on the thermal conductivity, we have also added a sentence in line 178-181 on page 8: "On the other hand, the variation in high P-T thermal conductivity of Fe-Si alloys is likely due to the P-T effects on electron-impurity scattering contribution to the conductivity (see Fig. S4 and Supplementary Information)." 3. Figure  4. Figure S4: This comment is somewhat related to the comment #2 above. Our results on Fe-Si alloys show that the temperature-dependent thermal conductivity likely changes with Si contents as well as applied pressure. As we have explained in line 137-145 on page 6, the co-existence of BCC phase with HCP phase is not a main contributing factor. To further clarify the P-T and Si effects on the thermal conductivity of Fe0.85Si0.15 alloy, we have revised Fig. S4 and Fig. 2. We added a red curve (which shows the averaged values of the high-T measurements at each given pressure) in Fig. 2 for the thermal conductivities (red symbols) to show the trend: the thermal conductivity of Fe0.85Si0.15 reaches a broad maximum around 80-100 GPa, followed by a decrease with higher pressure to 144 GPa. As the reviewer pointed out, such pressure evolution at high temperatures is complementarily revealed in Fig. S4, where we have also added the error bars to illustrate the measurement uncertainty. We explain the pressure-temperature dependence as follows and added a paragraph in the new section "Effects of pressure, temperature, and Si alloying on the thermal conductivity of Fe-Si alloys" in Supplementary Information: Thermal conductivity of Fe or Fe-rich alloy is dominated by the electrical conductivity.
The electrical conductivity (σ) and resistivity (inverse of conductivity) are strong functions of T (1/σ is proportional to T). However, the temperature dependence of thermal conductivity should be weaker as it can be determined via the Wiedemann-Franz (WF) law k=L×σ×T, where k is the thermal conductivity and σ the electrical conductivity, and L the Lorenz number. That is why a small change in the T dependence of resistivity with pressure would result in a change of the T dependence of thermal conductivity, which can increase or decrease with T (Fig. S4). This T dependence of thermal conductivity may also vary with the Si composition, making k decreasing with T (as in Konopkova et al., 2016) or increasing with T (as in this work at 106 GPa, Fig. S4). Please also refer to our responses in the comment #2 above regarding the effect of temperature on the thermal conductivity of Fe-Si alloys. Figure  from the probe side of the sample. The reviewer is correct that "the top side is the pulsed side and the bottom side is the opposite side of the sample" in our original submitted manuscript. To avoid confusion, we have changed the "top side" to "pulsed side" and "bottom side" to "probe side" throughout the revised manuscript.

5.
6. Figure S6:  has been provided in the original version of the manuscript. In the revised Fig. S6 caption, we have clarified the meaning of the symbols: Dark yellow (brown) stars are the P-T conditions of our high-temperature TH experiments collected from the pulsed (probe) side of the sample. These symbols represent the range of sample temperature variation measured by radiative temperature measurements from pulsed (probe) side of the sample.
Thus the original "bottom temperature" meant the temperature measured at the probe side of the sample, represented by the brown stars. Similar to the symbol issues in Fig. S5 in the comment #5, to avoid confusion, we have changed the "top temperature" to "pulsed side" and "bottom temperature" to "probe side" in the revised Fig. S6 Konopkova et al. (2016) at similar high P-T conditions (Fig. S1-3); the thermal conductivity difference between Fe0.85Si0.15 and pure Fe is well beyond the measurement errors labeled in the Fig. 2, presenting a solid evidence to support our conclusion.  Figure 2). 4. Figure S4 shows that the thermal conductivity of Fe0.85Fe0.15 increases with increasing temperature at pressures of 47~106 GPa, whereas it is independent from temperature at higher pressures. The author need to argue the mechanism to cause these phenomena.

The thermal conductivity of Fe and Fe-Si alloys increases by adding percent orders
This comment is similar to the comments #2 and #4 by the Reviewer #2 above. We explain it again as follows and added a paragraph in the new section "Effects of pressure, temperature, and Si alloying on the thermal conductivity of Fe-Si alloys" in Supplementary Information: Thermal conductivity of metals is dominated by the electrical conductivity. The electrical conductivity (σ) and resistivity (inverse of conductivity) are strong functions of T (1/σ is proportional to T). However, the temperature dependence of thermal conductivity should be weaker as it can be determined via the Wiedemann-Franz (WF) law k=L×σ×T, where k is the thermal conductivity and σ the electrical conductivity, and L the Lorenz number. That is why a small change in the T dependence of resistivity with pressure would result in a change of the T dependence of thermal conductivity, which can increase or decrease with T (Fig. S4). This T dependence of thermal conductivity may vary with the Si composition, making k decreasing with T (as in Konopkova et al 2016) or increasing (as in this work at 106 GPa, Fig. S4). However, we note that, given the Si alloying effect, it is expected that the Fe-Si alloys would have weaker temperature dependences than the pure hcp Fe, since the presence of impurities will enhance the scattering of carries (phonons and electrons) during their transport of energy. This qualitative behavior is clearly indicated in our data of Figs. 1 and 2. (Please also refer to our responses to comments #2 and #4 by the Reviewer #2 above.) To add further comment on the effect of temperature on the thermal conductivity, we have also added a sentence in line 178-181 on page 8: "On the other hand, the variation in high P-T thermal conductivity of Fe-Si alloys is likely due to the P-T effects on electron-impurity scattering contribution to the conductivity (see Fig. S4 and Supplementary Information)." 5. The outer core temperatures are more than 4000 K. The present temperature range is insufficient for arguing the core conductivity, because the temperature effect on thermal conductivity does not seem simple as argued above.
We hope the reviewer can understand that thermal conductivity measurements at simultaneous high pressure-temperature conditions are extremely challenging. Though the temperature conditions in our measurement is only up to 3300 K, our results have clearly demonstrated that the thermal conductivity of potential relevant composition in the outer core, Fe0.85Si0.15, is much lower than previously thought and pure Fe. This key conclusion is expected to remain valid at higher temperatures of about 4000 K because the temperature dependence of thermal conductivity is expected to be small at the temperature conditions of the outer core (the results at 132 and 144 GPa of Fig. S4).
6. I doubt that argument of the core thermal history based on thermal conductivity of core materials is essential. The core has cooled due to heat release across the CMB. If there were no heat production in the mantle, the released heat should be equal to the total surface heat flow. Thus, the core cooling rate should be controlled by the surface heat flow of the Earth not by the heat transfer in the core. This argument is strengthened by the fact that the thermal conductivity of rocks is much lower than that of Fe alloys.
Therefore, it is the thermal-boundary-layer structure of the bottom of the Earth that controls the amounts of heat released from the core to the mantle. The amounts of heat transfer within the core should be adjusted by verbosity of thermal convection in the outer core. I know numerous studies argued the cooling history of core based on Fe and Fe-alloy thermal conductivity as shown in Table 1 and 2, but I think that they are useless papers.
We are afraid that the reviewer's point of view about the relationship between the core thermal conductivity and thermal history is not precise. The reviewer is partially right, in principle, but wrong in the conclusions s/he gets from it. It is right that the mantle controls how much heat goes out of the core. However, even without heat production in the mantle, the CMB heat flow would not match the surface heat flow, unless the mantle is in steady state (i.e., not cooling down), which it isn't. Since the production of radiogenic heating in the mantle and the mantle cooling rate are far from being perfectly pinned down (and even more so for ancient eras), it is useful to find additional constraints on the problem. The persistence of the geomagnetic field for at least 3.5 Gyr is such a constraint that places minimum bounds on the CMB heat flow, and these bounds depend on the thermal conductivity of the core. The CMB heat flow could still be higher than these bounds and this point was made clear in the paper. The problem with the high conductivity values is that the minimum bounds that result from them are almost too high to be acceptable, although we can make it work, e.g. with extraction of light elements from the top. The low values proposed by our paper remove that difficulty. Therefore, determination of thermal conductivity of core Fe alloy at high P-T provides additional constraints on the thermal state and geodynamo of the planet.
Much of the information on the subject has been explained in the text in line 35-53.  table to Table S4 for the high-temperature measurements.
Following this recommendation, we have added more detailed descriptions of the high P-T experiments in the section "Thermal conductivity measurements at high pressure and high temperature" in Methods. We have also added a new table (Table S5) to list the parameters we used to analyze the data in the new Fig. S9 (originally Fig. S3). Figure S3 is not used anywhere in the text.

Page 28. I am afraid that
We are thankful to the reviewer for pointing out this omission. We have substantially extended the description of the high P-T experiments, and now this figure is cited. The new Fig. S9 (originally Fig. S3) shows an example data for the high temperature measurements: temperature evolution of Fe0.85Si0.15 foils at 121 GPa during flash heating at high initial temperature.
12. Page 32. The structures of these tables are confusing. One row has six columns and each of three columns describes one measurement. Even though space-consuming, each row should describe only one measurement.
We have added the borderlines to separate the pressure, temperature, and thermal conductivity for each of the measurements in Table S1-S3. We thank the reviewer for pointing this out.
The study of thermal conductivity of iron alloy is an important topic for the thermal evolution of the planet and the generation of a magnetic field. The authors' experiment point to much lower thermal conductivity for relevant alloys, compared with previous theoretical calculations or experiments based on measurement of electrical resistance.
The authors have done a good job responding to the reviewers comments. The experiments are carefully done and well explained. This will certainly help other researchers attempt to reproduce these results and consider other possible alloys. Some of the specific predictions likely depend on the choice of alloy elements, but the authors choice is reasonable and the conclusions they draw appear to be appropriate for their specific choice.
I think the paper can published in its current form. It will be of wide interest and generate consider (constructive) debate.
Reviewer #2 (Remarks to the Author): This manuscript provides the new data on thermal conductivity of Fe-Si alloys at high pressure and temperature, and showed the thermal conductivity of the alloy is very low compared to pure hcp-iron. I have checked carefully the revised manuscript. The authors addressed the reviewers' comments and the manuscript was improved significantly. This low thermal conductivity of the Fe-Si core provides very important implications for older formation of the inner core, energy source of geodynamo, early initiation of the Earth's magnetic field, and small heat flow at CMB. The results and implications are very significant, therefore this manuscript should be published in this journal.

Eiji Ohtani
Reviewer #3 (Remarks to the Author): This is the second review of this paper. Hence, I do not think that any preface is necessary in this review.
Although the authors claimed that the thermal conductivity difference between Fe0.85Si0.15 and pure Fe is well beyond the measurement errors, I disagree to their idea. The values obtained by the present method have actual uncertainties larger than a factor 2. This fact is suggested by the data points at pressures between 100 and 110 GPa, which range from 28 to 60 W m-1 K-1. If the error bars are taken into account, the actual uncertainty is a factor 4 (20 ~80 W m-1 K-1). Although low conductivity values (16-30 W m-1 K-1) were obtained at pressures above 120 GPa, these data should have the same level of errors. This should be also the case with Konopkova et al.'s study. A reasonable conclusion of the present study is therefore "the conductivity of Fe0.85Si0.15 is indistinguishable from that of pure Fe within the errors". Although I appreciate the authors' tremendous efforts for this pioneering study, the robustness should not be exaggerated. If the decrease in conductivity above 100 GPa of Fe0.85Si0.15 given in this study were the case, the conductivity of pure Fe given by Konopkova et al. should suggest that the conductivity decreases by a factor of 2 with increasing pressure from 40 to 70 GPa, and suddenly increase by a factor of 2.5 from 70 to 90 GPa, and then remain constant at higher pressures. These arguments based on Konopkova et al.' data are clearly too much, and therefore the authors conclusion regarding the difference between pure Fe and Fe0.85Si0.15 is also too much. I am afraid that high-pressure experimentalists tend to exaggerate the robustness of their studies. If a data set with the same quality is obtained by other methods than high-pressure experiments, no one would say that it is robust. I think that the present data set is worthwhile to publish in some journal. However, any geophysical argument is too early. The present study should be published at some journal of mineralogy such as American Mineralogist and Physics and Chemistry of Minerals. The authors should argue the thermal history of the Earth's core after being able to obtain that are more robust. For these reasons, I do not recommend this paper for publication at Nature Communications.
We thank you and the three reviewers very much for helpful comments. In particular, we appreciate that the final reports by Referee # 1 and # 2 are both very positive. In what follows, the remaining concerns and comments by Referee # 3 are in italics and our point-to-point responses are in blue. Changes in our revised manuscript are also highlighted in blue with tracked changes. Following your suggestion, we have toned down our data interpretations and made clear to the readers why the literature data that we took for comparison are rather scattered within a specific pressure range. Please see our revisions in the section "Thermal conductivity at high pressure-temperature conditions" (line 132-138 on page 6) of "Results" in the revised manuscript.

Report of Referee # 3
Although the authors claimed that the thermal conductivity difference between Fe0.85Si0.15 and pure Fe is well beyond the measurement errors, I disagree to their idea. The values obtained by the present method have actual uncertainties larger than a factor 2. This fact is suggested by the data points at pressures between 100 and 110 GPa, which range from 28 to 60 W m -1 K -1 . If the error bars are taken into account, the actual uncertainty is a factor 4 (20 ~80 W m -1 K -1 ). We should note that the data at 100-110 GPa the reviewer referred to actually have the largest uncertainties in our measurements, so this is an extreme case in this study. In comparison, the Fe 0.85 Si 0.15 data above 120 GPa have relatively small uncertainties. The data uncertainties presented in Fig. 2 are somewhat scattered and reflect the current status of thermal conductivity measurements at extreme pressure-temperature conditions. Even with these uncertainties considered, the overall data trend clearly show that thermal conductivity of hcp-Fe 0.85 Si 0.15 alloy is lower than that of pure hcp-Fe at relevant pressures of the core such as 132 and 144 GPa. Therefore, our geophysical implications on lower thermal conductivity of Fe-Si alloy in Earth's core are supported by the data.
Although low conductivity values (16-30 W m-1 K-1) were obtained at pressures above 120 GPa, these data should have the same level of errors. This should be also the case with Konopkova et al.'s study.
The data uncertainties in this study are derived from standard error propagation procedures, including averages from multiple measurements and error propagations in data modelling. Thus, the uncertainties also depend on the experimental data quality and are not necessarily the same in each experimental run.