Presentation on theme: "Ferromagnetism in HPHT C 60 polymers: magnetic carbon or magnetic impurities? Alexandr Talyzin, ( Department of Physics, Umea University, Sweden )"— Presentation transcript:
Ferromagnetism in HPHT C 60 polymers: magnetic carbon or magnetic impurities? Alexandr Talyzin, ( Department of Physics, Umea University, Sweden )
A.Talyzin, Umea University, Sweden 1. History of magnetic carbon. Questions? 1996, 1998, 2000- Synthesis of 3 sets of samples. Metallic conductivity in high pressure polymer of C 60 (Makarova et al. Synth. Met. 121, (2001).) 2001- Pressure polymerized C 60 is intrinsically ferromagnetic (as occasionally discovered) Makarova et al., Nature 413, 716-718, (2001) 2002-2004 – four experimental papers confirming ferromagnetism of high pressure polymers of C 60 (from two groups) 2002-2006- about 15 theoretical papers aimed to explain high temperature ferromagnetism of C 60 polymers. 2005- Ferrocarbon: EU project aimed on study of Magnetic Carbons. August 2005 – Corrigendum to original paper on magnetic carbon is published in Nature 30 March 2006 – The Nature paper is retracted by 7 authors out of 9 citing results presented below. Two authors (T.Makarova and P.Scharff have not joined retraction).
Polymeric C 60 - magnetic carbon? Figure by T.Makarova,2003- 2005. Summary of discovery ( Makarova et al, Nature,2001 ), five main arguments: 1. Synthesis temperatures close to the point of C 60 collapse. P=6 GPa, T=1025-1050K 2.Ferromagnetic with Curie T ~500K 3. Reproducible: 3 sets of samples, one of them specially synthesized were reported in original Nature paper. 4. Ferromagnetism disappears after thermal depolymerization of samples (700K) 5. 22 ppm level for all ferromagnetic impurities (0.003 emu/g) (30 fold less compared to measured magnetization)
No confirmation ( Han et al, Phys.Rev.B, 2005 ) A.Talyzin, Umea University, Sweden Over 20 samples synthesized at 3.5 GPa. Structure and samples and synthesis conditions are very close to those from Nature samples. X-ray diffraction from one set of samples synthesized in 2004 and studied by SQUID in Germany (Prof.Esquinazi group) All samples were diamagnetic (including polymers) Magnetization traces after subtraction of diamagnetic contribution: 0.0003 emu/g 0.1 emu/g in Nature, 2001 Why …?
Ferromagnetism disappears after thermal depolymerisation at 700K (Nature, 2001) DSC (A) R-phase; (B) Magnetic carbon. with Curie T>800K (Korobov et al, Chem.Phys.Lett, 2003) Nature: magnetization reversible after heating at 640K for 2 hours Adv.Mat, 2002: Magnetization is reversible After heating at 800K for 16 hours The same sample (E17) depolymerized at 200 0 C (473K) in 2001 ( Makarova et all, Carbon ). In Nature (2001) depolymerization was reported below 700K Depolymerisation of C 60 polymers occurs at 500-560K (1998). (including photopolymers) Ferromagnetism can not be assigned to polymeric structure.
How to prove existance of new ferromagnetic material? M s =0.1 emu/g Intrinsic ferromagnetism of bananas? Fe Ms=220 emu/g ~450 ppm (0.045%) of Fe- 0.1 emu/g Any claim of new ferromagnetic material MUST be provided with impurity concentrations- Fe, Ni, Co Ideally, analyses should be done on the same sample. Analysis of starting chemicals is NOT sufficient-final sample must be analyzed
Impurity analysis (Makarova et al, Nature 2001). A.Talyzin, Umea University, Sweden -We have paid great attention to chemical analysis of the pristine material as well as of the polymerized phase. The total amount of magnetic (Fe,Ni, Co) impurities is 22 ppm in the pristine phase. -Sample cited in Nature: 3 mg. Impurity analysis mentioned in the paper (atomic absorption spectrometry) required 20 mg. Contamination introduced during synthesis was not taken into account. Toroid apparatus used for synthesis of samples (Prof. Davydov group,Troitsk, Russia)
Independent confirmations. 1) 2.5 GPa- Makarova et al (Synth.metals, 2003). No impurity analysis was done, no magnetic data shown, M s = 0.01 emu/g (~50 g/g Fe) 2) 6 GPa- Narozhnyj et al (2003), proceedings, 22 ppm in pristine C60 (with reference to Nature paper). Curie T above 800K, not determined. 3) 9 GPa- Wood et al, (2002). Impurity analysis not presented at all. Only one sample shown to be ferromagnetic. 4) 9 GPa- Chan et al,2005 (the same group as 3) Hydrogen amount determined but not Fe, Ni,Co. None of the papers provided satisfactory Impurity analysis.
22 ppm (mg/g) for all magnetic impurities? No, 400 ppm. A.Talyzin, Umea University, Sweden Re-analysis of Nature samples by R.Hohne and P.Esquinazi (Adv.Mat., 2002) -Iron contamination measured by PIXE (30 m depth): 200- 400 g/g, about 20 times higher than previously reported (enough to explain ferromagnetism) Only 3 samples were studied by SQUID before submission to Nature (one split on two pieces) (figure from Adv.Mat., 2002)
Reproducibility and history of samples. All samples used for Nature paper were NOT INTENDED (as stated by Prof. V.Davydov) for studies of ferromagnetism. -All samples were synthesized prior discovery of ferromagnetism, specially synthesized set (2000) as well. First observation of ferromagnetism in March 2001 (Prof.P.Esquinazi). -Samples were split in 1996,1998,2000 by metallic tools (stated in published papers) and repeatedly touched by unprotected magnet in 2001-2004. -No precautions against contamination was taken during synthesis, several years of storage and handling of samples. Video from supplementary materials available on line ( Makarova et al, Nature, 2001 ) The same treatment until 2004 was given to every new sample…
Amounts of iron in magnetic carbon A.Talyzin, Umea University, Sweden Figure from Han et al, Carbon 2003 Figure from Spemann et al, (2003), conference proceedings Particles of tens micrometer size. ImageJ software shows 2-3% of surface are covered by subsurface particles (counting only sharp ones). (~30 m depth of method). 22 ppm of Fe?? Averaging from PIXE: 2-3 m spots, 5 points. 1.370, 100, 200, 16.000 and 100 (average 482 g/g) Hundreds of points required!
Conclusion: no impurity analysis-no intrinsic ferromagnetism No evidence for intrinsic ferromagnetism in HPHT polymers of C 60. Additional arguments: 1.Ferromagnetism observed for samples synthesized in short interval of temperatures (1025-1050K)? 2.How to explain 500K Curie T?
How impurity could explain Curie T of 500K? A.Talyzin, Umea University, Sweden Fe 3 C is ferromagnet with Curie Temperature 480-500K Fe, Nb, W, V are known to induce collapse of C 60 with formation of carbides in thin films already at 400K. Model experiments: Fe+C 60 at HPHT conditions Two samples of C 60 : mixed with 10% (mass) of Fe and 3% of Fe. Samples subjected to the same P-T treatment as C 60 used for preparation of magnetic carbon. Uniform size of Fe particles: 2-3 m
Reaction of Fe with C 60 at 2.5 GPa and 1040K A.Talyzin, Umea University, Sweden Iron peaks disappear after high pressure high temperature treatment. Excess of C 60 transforms into polymeric phase. More deteail XRD: iron transformed into Cementite, Fe 3 C. Curie T (Fe 3 C)= 480-500K Fe 2 O 3 +C Fe 3 O 4 +C Fe+C Fe 3 C
Ferromagnetism in C 60 polymer/Fe 3 C mixture (Prof. P.Esquinazi group, University of Leipzig) A.Talyzin, Umea University, Sweden Pristine C 60 /10 % Fe mixture HPHT treated with 10% Fe HPHT treated with 3% Fe Saturation magnetization of Fe 3 C at (RT) M s =128 emu/g. Expected magnetic moments at saturation due to Fe 3 C were calculated from known Fe concentrations and masses of samples. Values calculated for both samples are in agreement with the measured data within experimental error. Iron do not induce ferromagnetism in in carbon
Comparing C 60 /Fe 3 C to magnetic carbon. A.Talyzin, Umea University, Sweden Pristine C 60 /10% Fe mixture Right axis: two samples of magnetic carbon: E17 ( Nature, 2001 ) and E16 ( Adv.Mat., 2002 ) Left axis: C 60 /Fe 3 C samples obtained at HPHT. Conclusion: magnetic carbon and C 60 /Fe 3 C show nearly identical magnetic properties. Rejected from Phys.Rev.B, Published in Eur.Phys. J B., 2007
Ferromagnetism is found reproducibly only for synthesis temperatures close to the point of C 60 collapse? A.Talyzin, Umea University, Sweden -Nature (2001)- 6 GPa, only at T=1025K-1050K (just below collapse point 1075K) -Review paper (Makarova, Semiconductors, 2004) -about one hundred degrees below point of C 60 collapse.. That will be 975K. Out of range. -Review by Makarova, JMMM, (2003)- ….with maximum at 1075K, exactly the point of collapse. -MFM paper (Han et al, Carbon, 2003) only one sample studied, the same as in Nature: synthesized at ….P=6 GPa and 1125K Above point of C60 collapse! -The same study published earlier (Nucl. Instr. and Meth. in Phys. Res. B 210 (2003) 531–536) : P=2.5 GPa and 1125K, this sample could not be Rhombohedral at this pressure.
Verified synthesis conditions for Nature samples (Corrigendum, 2005). A.Talyzin, Umea University, Sweden Stated in the paper: 6 GPa and 1025-1050K, ferromagnetism only in this narrow interval of temperatures, 5 samples ferromagnetic from 3 different sets. Only three samples were studied by SQUID (Prof. Esquinazi, Germany) prior to paper submission. Only two of them were ferromagnetic: - Nature sample 6 GPa and 975K lowest temperature from all set. -2.5 GPa and 1125K (1998) both P and T out of reported range. - 6 GPa and 1025K (2000) Not ferromagnetic. Corrigendum published in Nature, 2005. Reports minor mistakes in synthesis conditions. Celsius to Kelvin calculation error, mistake with sample labels.
Implications of new synthesis conditions A.Talyzin, Umea University, Sweden -Only one ferromagnetic sample was true Rhombohedral polymer, synthesis temperature 975K was lowest from all samples. Narrow interval of temperatures do not exist. -Second sample (2.5 GPa and 1125K) was not polymeric and not fullerene: C 60 collapses at these temperatures. -Magnetic properties of these two samples were identical: ferromagnetism could not be connected to fullerenes! -Curie T of 500K was EVER observed only for these two samples, only one of them true Rh polymer. Next set of samples showed Curie T above 800K (Narozhnyj, 2003).
Summary and Retraction (29 March 2006, Nature) A.Talyzin, Umea University, Sweden..T.L.Makarova and P.Scharff decline to sign this retraction because they do not believe that the earlier results, supported in subsequent studies, are totally invalidated by these findings…. Claim is based on unpublished data for both magnetization measurements and impurity analysis. Retraction (2006) signed by : B. Sundqvist, R. Hohne, P. Esquinazi, Y. Kopelevich, V. Davydov, L. S. Kashevarova & A. V. Rakhmanina Ferromagnetism of high pressure polymers of C 60 : - not connected to polymeric structure - not connected to C 60 - not reproduced in new experiments - level of magnetization can be explained by impurities without suggesting magnetic carbon. - Curie T of ~500K was ever observed only for two samples and can be explained by Fe 3 C
Magnetic carbon in meteorite? (Coey et al, Nature, 2002) A.Talyzin, Umea University, Sweden Magnetisation measured from graphite nodule reported to be 30% higher compared to magnetisation calculated from mineral composition. Curie Temperature of magnetic graphite : ~500K Interpretation: iron induces ferromagnetism in graphite. Fe 3 C is actually named as mineral composing the meteorite. Cohenite (typical mineral of metheorites): Fe (Ni,Co) 3 C- antiferromagnetic or ferromagnetic, not counted in the calculations. Cementite: (also known in meteorites) Fe 3 C- ferromagnetic stronger than magnetite.
Cigarette ashes and magnetic meteorites A.Talyzin, Umea University, Sweden Thermal demagnetisation of ashes from Bulgarian cigarettes Shipka (N.Jordanova, Journal of Magnetism and Magnetic Materials, 2006). Camel and Marlboro also studied… Fe 2 O 3 - Fe 3 O 4 - Fe 3 C reactions occur at temperatures higher than 1020K in process of cigarette burning Fe 3 C Fe3O4 Thermal demagnetisation of Magnetic meteorite, J.M.D. Coey et al., Nature (2002) Fe must have reacted with carbon at temperatures of meteorite falling and impuct.
Metallic conductivity in high pressure polymers of C 60 : result of mistakes in Celsius to Kelvin calculations Makarova, T. L. et al. Anisotropic metallic properties of highly-oriented C 60 polymer. Synth. Met. 121, (2001). Makarova, T. L. et al. Electrical properties of two-dimensional fullerene matrices. Carbon 39, (2001). Okotrub et al, J.Chem.Phys. (2001) –model proposed to explain metallic Rh C 60. Makarova (JMMM), 2004 1075K 1025K 975K Makarova, (2002) 1025 K 1050 K 1075K Makarova, (Mol.Mat) 2000 1048K 1073K 1123K Point of C 60 collapse: 1075K (Nature)
More carbon magnets? A.Talyzin, Umea University, Sweden - Proton irradiated graphite (Prof. P. Esquinazi group). Impurity analysis presented, interesting subject to study. -Carbon nanofoam (A.V.Rode, Phys.Rev.B, 2004 ). Impurity analysis presented. Ferromagnetism below 90K -Chemically etched graphite (A.V.Mombru, PRB, 2005). Impurity analysis presented. Results not confirmed (J.M.D.Coey group, 2006), large Fe particles found in their samples in Umea (K. Han) - Talapatra et al, Phys.Rev.Lett., 2005: ferromagnetism in nitrogen irradiated Nanodiamonds. Commercial detonation nanodiamonds are always contaminated with metallic alloy catalyst. Impurity analysis not presented.
Questions ? Acknowledgements: - Prof. B.Sundqvist, (Umea university). - A.Dzwilewski, samples synthesis (Umea University, Sweden) - Prof. L.Dubrovinsky, facilities for synthesis (Bayreuth University, Germany) - Prof. P. Esquinazi and A. Setzer and (University of Leipzig, Germany). SQUID magnetometry measurements. -Special thanks to Prof. P. Esquinazi for providing us with all available raw data related to C 60 magnetic carbon and many discussions. More details in recent papers: 1.A. Talyzin, A. Dzwilewski, L. Dubrovinsky, A. Setzer, and P. Esquinazi, Structural and Magnetic properties of polymerized C 60 with Fe. Eu. Phys. J.l B, 2007. 2.A. Talyzin and A. Dzwilewski,J.Nanoscience and Nanotechnology, Review, April 2007
Increased stability of magnetic carbon to depolymerization? A.Talyzin, Umea University, Sweden Depolymerisation of ALL C 60 polymers occurs below 600K (including photopolymers). T.Makarova, Semiconductors, 2004 XRD shown in figure were taken from two different samples, both are not from Rh polymer! Heated at 800K (2.5 GPa, 1125K) Pristine polymer (2.5 GPa, 1050K) Makarova (Semiconductors, 2004): One of the samples lost only 2% of magnetization, and the X-ray diffraction patterns remained unchanged after annealing and corresponded as before to the rhombohedral phase of polymerized C 60. This sample (E17) was destroyed during heating due to oxygen leak (Prof. Esquinazi) The same sample (E17) depolymerized at 200 0 C in 2001 (Makarova et all, Carbon). In Nature (2001) depolymerization was reported below 700K