Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The structural and magnetic properties of some ternary alloys of iron-aluminum-carbon Bell, Lawrence Gerald 1955

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1955_A7 B3 S9.pdf [ 1.82MB ]
Metadata
JSON: 831-1.0106335.json
JSON-LD: 831-1.0106335-ld.json
RDF/XML (Pretty): 831-1.0106335-rdf.xml
RDF/JSON: 831-1.0106335-rdf.json
Turtle: 831-1.0106335-turtle.txt
N-Triples: 831-1.0106335-rdf-ntriples.txt
Original Record: 831-1.0106335-source.json
Full Text
831-1.0106335-fulltext.txt
Citation
831-1.0106335.ris

Full Text

THE STRUCTURAL AND MAGNETIC PROPERTIES OF SOME TERNARY ALLOYS OF IRON-ALUMINUM-CARBON by LAWRENCE GERALD BELL A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in the Department of MININGAND METALLURGY We accept this thesis as conforming to the standard required from candidates for the degree of MASTER OF APPLIED SCIENCE. Members of the Department of Mining and Metallurgy. THE UNIVERSITY OF BRITISH COLUMBIA. October, 1955. ABSTRACT Alloys i n and near the £. single phase region found by Morral In the iron-aluminum-carbon system were prepared and their ferromagnetic and structural properties determined„ The t phase has been described as chiefly Fe3Al plus 15 atomic percent carbon with the possibility of varying the Fe:Al ratio from approximately 2 to 3.5. The face-centred cubic £ phase was found to be highly ordered with iron at the face centre positions, aluminum at the cube corners and carbon in the body-centred position. The la t t i c e parameter varies with carbon content from 3.73 to 3.76 A 0. Keeping the amount of carbon constant at 1^ .6 atomic percent the alloys with the Fe:Al ratio less than 3 have tbeir saturation magnet-ization increasing with increasing iron by an amount corresponding to the estimated increase of 5.4 Bohr magnetons per iron atom. However, for values of the ratio greater than 3 the magnetization is decreased with increasing iron. The amount of this decrease i s not certain but i t i s thought to be of the order of 8 Bohr magnetons per iron atom i n excess of that required to give FeiAl equal to 3. Increasing the carbon content with the Fe:Al ratio kept at 2.9 * .1 also decreases the magnetization by an amount corresponding to an estimated 8.5 Bohr magnetons per carbon atom. ACKNOWLEDGEMENT The author is grateful for financial aid in the form of a National Council Bursary during the summer of 1954 and the university year 19/*5~55« The work was done with the help of funds provided.by the Defence Research Board under Research Grant 261. The Department of Mines and Technical surveys obtained assay results for.ali three constituents in the alloys and many carbon assays were also done by the Vancouver Steel Company. The author is grateful for the assistance of the staff of the Department of Mining and Metallurgy. Special thanks are extended to Professor F. A. Forward for providing the facilities for the work and also to Dr. H. P. Myers, the director of this research, and R. G. Butters for technical advice and encouragement. TABLE OF CONTENTS Page I. INTRODUCTION 1 II. PREVIOUS WORK 1. Isomorphous alloys containing managanese . . . . . . . 3 a. magnetic properties b. structural properties 2. The Fe-Al-C system 6 III. EXPERIMENTAL PROCEDURE 1. Preparation of the alloys 7 2. Heat Treatments 7 3. Metallography 8 4. Structure and Lattice Parameter measurements 8 5. Magnetic Properties 10 IV. RESULTS 1. Metallography 12 2. Structure and Lattice Parameter measurements 12 3. Magnetic Measurements 19 V. DISCUSSION OF RESULTS AND CONCLUSIONS 31 VI. APPENDICES 1. Particular difficulties encountered in preparation 36 of the alloys 2. Calculations of change in Bohr magneton number . . . . 37 VII. BIBLIOGRAPHY 39 ILLUSTRATIONS Page 1* The proposed structure of Mn^AlC and Mn^ ZnC. (Butters and Myers, 1955) 4 2. The iron-rich corner of the iron-aluminum-carbon system at 1000°C as proposed by Morral (1934) 4 3. The variation of the Curie temperature of the phase with aluminum content. (Snoek, 1945) 6 4. The proposed section of the iron-aluminum-carbon system in the region of the c phase at 1050°C. 14 5. Variation of the lattice parameter of the 6 phase with carbon content . . . . . 16 6. Variation of the lattice parameter of the 6 phase with Fe:Al ratio 17 7. Graph of Curie temperature, 0e, versus atomic percent carbon for alloys of Fe-Al-C 20 8. Variation of Curie temperature, 6e, with Fe:Al ratio for alloys of Fe-Al-C 21 9« Graph of saturation magnetization, S0, versus atomic percent carbon for alloys of Fe-Al-C with Fe:Al ratio approximately 2.9 22 10. Variation of saturation magnetization, G0, with FetAl ratio for alloys of Fe-Al-C . 23 11. Variation of the spontaneous magnetization, C , with tempera-ture for some of the alloys with different carbon content and Fe:Al approximately 2.9 .. . ." ". 24 ILLUSTRATIONS (continued) Page 12. Variation of the spontaneous magnetization, $ , with temperature for various alloys with different Fe:Al ratios and almost con-stant carbon content 25 13. Variation of Bohr magneton number versus atomic carbon for alloys of Fe-Al-C 26 14. Variation of Bohr magneton number with Fe:Al ratio for alloys of Fe-Al-C . . . . . . . . 27 15. Graph of Bohr magneton number versus Fe:Al ratio. A l l points have been corrected to the expected value for alloys with the same Fe:Al ratio and 14.61 atomic percent carbon 28 16. Magnetic moment per unit mass divided by saturation moment, against absolute temperature divided by the ferromagnetic" Curie point T/e, 29 TABLES Page 1. Composition in atomic percent of the alloys studied and metallographic observations. . . . . . . . . . 13 2. Properties of the 6 phase in alloys of Fe-Al-C . 15 3. , Results of X-ray line intensity measurements 18 1. THE STRUCTURAL AND MAGNETIC PROPERTIES OF SOME TERNARY ALLOYS OF IRON-ALUMINUM-CARBON I. INTRODUCTION Interest i n the magnetic properties of the alloys i n the region of composition Fe 6 0Al 2oC2o arose from a study of the alloys Mn60Al2c>C2o a n c* Mn60ik?2pC8o» These alloys possess interesting structural and magnetic properties which are not as yet completely understood. An investigation of isomorphpus alloys having preferably only one constituent different from the manganese alloys may provide useful data for a proper understanding of their behaviour. It i s for this reason alone an investigation of the iron-aluminum-carbon alloys was started. Very l i t t l e work has been done on the iron-aluminum-carbon system i n the region of interest and especially on the magnetic properties of the alloys. F. R. Morral proposed a phase diagram for the iron-rich corner of the system but his data in the region of the interesting phase, which'he designated by E , was very limited. Some interest was shown in these alloys by J. L. Snoek from the point of view of their being potential permanent magnet material. Alloys of this composition were found, however, to be unsatisfactory although those with lower aluminum and carbon content were useful. In the present work alloys have been made in the region of composition of the single phase £ i n an attempt to f i x more exactly the limits of s t a b i l i t y . Metallographic methods were used i n this pursuit. Alloys were also prepared having different Fe:Al ratios with the object of 2 maintaining the carbon content constant at 15 atomic percent and their properties determined. The purpose of this was to observe the variation in properties with aluminum content in analogy with work done on the manganese-aluminum-rcarbon and manganese zinc-carbon systems. In order to observe the change in properties with carbon content an attempt was made to obtain alloys with Fe:Al slightly less than 3 and with different carbon concentrations. II. PREVIOUS WORK 3 Isomorcihoiis. Alloys Containing Manganese. Previous work has shown that several ternary alloys containing manganese and carbon are single phase and exhibit ferromagnetism. The third constituent has a high valency and a positive size factor with respect to manganese; for example, aluminum, zinc, tin and indium have been used successfully in this capacity. The magnetic properties and structure of those alloys containing aluminum and zinc have been investigated by Butters and Myers (1955 A and B). The optimum compositions of these alloys are Mn3AlG and M^ZnC and they have been found to possess highly ordered face-centred cubic structures even after drastic quenching, as for example in the c h i l l cast state. In this ordered state the manganese atoms are at the face centres, the aluminum or zinc atoms at the cube corners and carbon at the body-centre position (see figure 1). The unusual magnetic properties of these alloys are of most importance here but a study of their stability ranges will also help in understanding the structure of the alloys. a. Magnetic properties. The alloy MnjAlC is strongly magnetic at low temperatures, having saturation moment at 0°K of 99.6 erg/gm oersted which is equivalent to a Bohr magneton number of 1.20 per manganese atom. The Curie temperature is 15°C. when measured in a magnetic field of 16,200 gauss. As the-MnsAl-ratio is increased above 3 with the carbon content held constant at 20 atomic percent the aluminum atoms are in effect -being replaced by manganese atoms and the Bohr magneton number is decreased an amount corresponding to 5 * 0.5 Bohr magnetons for each manganese atoms in excess of 60 atomic percent. A possible interpretation of this is that the excess manganese atoms give rise to a moment antiparallel to that of the manganese atoms at the face-centre positions. The paramagnetic o o o Figure 1 The proposed structure of M113AIC and Mn3ZnC (Butters and Myers, 1955) 0 Aluminum or zinc O Manganese 9 Carbon g 14 -1-1 * V o u £ 0 1 2 3 4 5 6 7 Wt. percent Carbon. Figure 2 The iron-rich corner of the iron-aluminwm-carbon system at 1000°C as proposed by Morral (1934). susceptibility for this alloy was found to vary with temperature in a manner different from a normal ferromagnetic in that the L/x v ersus T graph has a pronounced curvature concave to the temperature axis. This is similar "to the observed characteristics of a ferrimagnetic (Neel, 1948) but a proper interpre-tation of the behaviour of the alloy M^AIC remains to be found. The alloy Ifc^ZnC exhibits a marked anomally at low temperatures, the saturation magnetization decreasing with decreasing temperature. A maximum in the magnetization occurs at -42°C which may be interpreted as a criti c a l temperature below which an ordering of the magnetic moments exists. This crit i c a l temperature is also associated with a continuous change in structure from face-centred cubic to face-centered tetragonal. This is a new and hitherto unobserved feature of magnetic behaviour and gives added interest to a study of the properties of isomorphous alloys. b. Structural properties. The ordering in alloys of the type Mn3XC is not expected to be of the usual superlattice type because of the stability of the structure with temperature. It may be that the cr i t i c a l ordering temperature is higher than the melting point, but because of the presence of carbon this possibility should perhaps be discarded in favour of a bonding involving the carbon atoms. The minimum ratio of Mn:Zn and Mn:Al in these alloys with 20atomic percent carbon is in both cases believed to be 3 whereas the maximum ratio is about 8. This implies that the aluminum or zinc cannot take up the face-centre positions but that the manganese may replace the aluminum on the corners. A possible simple explanation of this is that the manganese atoms are -held at the face-centres by a strong bonding to the carbon . atom. 6. The Fe-Al-C System The iron-rich corner of the ne^aluminum carbon system at 1000°C as proposed by Morral (1934) is shown in figure 2. He found the £ phase to be a highly ordered face-i-centred cubic structure with lattice parameter varying between 3.73 and 3.77 A°. Calculations of line intensities for aluminum at the cube corners and iron at the face centres agreed with those observed on an X-ray photogram. The magnetic properties of the g phase have not been studied carefully; i t was considered ferromagnetic by Morral, and the Curie temperatures of some alloys were measured by J. L. Snoek (1947). From the large variation in the Curie points as determined by him (figure 3), he concluded that the £ phase is not a uniform compound but may have various compositions. 300 200 100 0 -100 7 8 9 10 ] .1 12 ] L3 14 1! Wt. £ Al. Figure 3 The variation of the Curie temperature T c of the € phase with aluminum content. (Snoek, 1945). I l l EXPERIMENTAL PROCEDURE 7. Preparation of the Alloys The materials used in this work were iron of 99.8 percent purity -either that donated by Plastic Metals Company or Westinghouse pure iron, aluminum of 99.99 percent purity donated by the Aluminum Company of Canada and graphite of spectroscopic grade. The alloys were prepared by high frequency induction heating in an argon atmosphere. Weighed amounts of iron and carbon making up the desired composition were placed in a high grade alumina crucible and were first heated to a red heat to allow degassing to occur. The furnace assembly was then flooded with purified argon, evacuated and once more filled with argon to almost atmos-pheric pressure. This procedure was carried out to prevent the loss of carbon and aluminum by oxidation. It was found that an unpredictable amount of carbon was lost during melting, in some cases up to 0.5 and even 1 percent of the total specimen weight. An attempt was made to compensate for this but it was not very successful since the amount lost was not constant. After melting and mixing the iron and carbon under argon atmosphere, the crucible was raised so that a piece of aluminum suspended on a thread was consumed by the melt. The total alloy was then heated to -well above the melting point to allow for good mixing, and then chill cast into a split brass mould. There was no apparent contamination of the melt by the mould but many difficulties were encountered. (See Appendix l ) . Heat Treatments. All alloys were given the heat treatment of homogenization for four days at 1050°C in an evacuated quartz tube, then quenching in water from this temperature. It was found that homogenization below 1000°C and in particular at 950°C resulted in the precipitation of the undesirable y phase in a l l alloys in the interesting range. From this i t was concluded that the E phase is unstable at temperatures below 1000°C and the heat treatment at 1050°C was adopted. A l l homogenized samples were found to have a small amount of secondary phase on the outside surface. This was probably caused by the loss of some aluminum by vapourization. The attempt to minimize this effect by using small quarts tubes and keeping them as f u l l of specimen as possible proved successful in most cases. (See Kidson, 1953)o For measurements a clean sample was taken from the centre of the homogenized specimen. Metallography Al l homogenized samples were mounted in lucite, polished, etched with 4 percent nitol or 1 percent HF and observed under the microscope. It was found that nitol did not attack the g phase but the cubic a phase (see figure 1) was etched to a lighter colour. HF deeply etched this a phase and also outlined the ferromagnetic y phase which appeared to nucleate along grain boundaries. No etchant was found to indicate the structure of the 6 phase although i t was found to be soluble in dilute HC1 and HNO3 solutions. Free carbon in the specimens was observed as l i t t l e nodules, termed rosets by some-authors, and also as flakes distributed throughout the white 6 phase. Structure and Lattice Parameter Measurements. The -structure and lattice parameters of the alloys were determined by X-ray diffraction techniques. The alloys were very brittle so that powders were -easily prepared and i t was found unnecessary to anneal the powders to get sharp diffraction lines. Debye-Sherrer powder photographs were taken and from these lattice parameters were calculated in the usual way. For structure deter-minations the intensity of diffraction lines was measured using a Geiger counter spectrometer. Line intensities were obtained by measurement of the line areas above the background level and these were compared with those calculated for the assumed perovskite structure with 100 percent order. In these calculations the temperature factor was omitted but the atomic •scattering factor was corrected for any depression in the region of the absorbtion edges (James, 1946). The formula for relative intensities is I 1 + cos22 9 • P • (F)* s i n 2 e cos e where 0 is the angle of diffraction, P is the multiplicity of contributing planes and (F) is the geometrical factor. The latter takes the values: ( f A 1 - f F e 4- f c ) 2 when Z.h2 is odd 2 2. and (fjQ - fp e - f c ) when £ h is even for the superlattice lines of the face-centred cubic structure ( 2 h 2 = 1, 2,. 5, 6, 8, 9, 13), and (3f F e + f A 1 + f c ) 2 when Zh 2:is odd and O fFe + f A l " f c ^ w h e n ^ ^  is even for the normal diffraction lines of the face-centred cubic, structure ( Z h 88 3, 4, 8, 11, 12). is the atomic scattering factor for atoms on the cube corner positions and fp e, f c are the same quantities for the face-centred and body-centred positions respectively. Alloys with the ratio Fe^Al different from 3 are assumed to have excess aluminum or iron take up positions in the other lattice at random so that f a n d fpe a r e average values. f c is as well a fraction of the atomic scattering factor for carbon since a l l of the body centre positions are not f i l l e d . 10. -Magnetic Properties For the magnetic measurements an external field of 16,200 oersted was obtained by means of an electromagnet having its poles shaped so that a uniform field gradient was produced. The force exerted on a specimen of mftss m and magnetic moment per unit mass (J , is given by F = m frH A Sucksmith ring balance having a deflection proportional to the force exerted on the sp^imen was used to compare this force with that exerted on a standard iron sample of known magnetization. The saturation moment may then be calculated without knowing 3 H since F l , S.i m l •„ F s esms K Measurements of ff were made over the range of temperature between 117°^ and the Curie point, which was in general less than 330°C. Temperatures below room temperature were obtained by allowing the sample- to slowly warm after being submersed in liquid oxygen. Above room temperature a furnace assembly with facilities for producing a vacuum around the specimen was used. From the values of ff and T,;the saturation magnetization at absolute zero was obtained by plotting T .. against ff to produce a straight line and extrapolating this to 0°K. Curie temperatures were taken as the extrapolated values of ff against T to ff * 0. An attempt was made to measure the paramagnetic behaviour of the alloys above the Curie temperature. This was done.using a similar but more sensitive ring balance- than that used in the ferromagnetic, range. About 130 milligrams of coarse powder were put in a tiny carbon crucible in the 11. form of a ring which was held on a quartz tube between the poles of the magnet• and heated i n vacuum. 12. IV. RESULTS Metallography Alloy compositions and the results of metallographic observations are given in table 1. These have been summarised in the phase diagram shown in figure k, in which the proposed secrtion of the ternary phase diagram at 1050°C is given, as well as the section at 1000*0 proposed by Morral. Structure and Lattice Parameter Measurements. The lattice parameter for the £ phase in most of the alloys is given in table 2. In figure 5 this quantity is plotted against the atomic percent of carbon and in figure 6 against the Fe:Al ratio. It is apparent that within the accuracy of these measurements the parameter changes little with the Fe:Al ratio, but an increasing amount of carbon increases the parameter from 3.730 A 0 at 9$ carbon to 3.755 A° at 16.3 percent carbon. The results of line intensity measurements as well as the calculated values for 100 percent order are given in table 3. The two alloys dealt with in this case are numbers 5 and IS with Fe:Al ratio of 2.87 and 3.22 respectively. Measurements were repeated six times on alloy 5 and twice on alloy 18. The average values of measured intensity agree with those calculated to within 10 percent for most lines, which is within the limits of expected error for such measurements. The large difference between the observed and calculated intensities for the lines with- index 1 may possibly be explained by the increase in specimen size at small angles. TABLE 1 Composition in atomic % of the alloys studied and metallographic observations. (Compositions by volume). Alloy Iron Aluminum Carbon % t % c 1 66.22 16.78 17.01 80 0 15 5 2 67.61 17.95 . 14.42 90 0 10 0 3 . 64,27 19.35 16.35 95 0 0 5 4 62.47 21.23 16.28 95 0 0 5 5 63.53 22.13 14.36 100 0 0 0 6 65.13 22.90 11.98 95 5 0 0 7 65.75 21.55 L2.71 95 5 0 0 8 62.84 23.32 13.86 100 0 0 0 10 62.55 23.88 13.56 100 0 0 0 11 63.74 23.88 12.37 12 61.82 24.74 13.43 99 1 0 0 13 61. IB 25.29 13.55 98 2 0 0 18 65.18 20.21 14.61 100 0 0 0 19 68,46 22,28 9.25 70 30 0 0 20 64.25 22.30 13.45 100 0 0 0 22 63.17 21.68 15.15 100 0 0 0 24 64.51 20.66; 14.83 100 0 0 0 15 20 Atomic f0 C -*• Figure 4 The proposed section of the iron-aluminum-carbon system in the region of the £ phase at 1050°C. (The dotted lines show the section at 1000°C proposed by Morral), TABLE 2 Properties of the £ phase in alloys of Fe-Al-C. Alloy #C. FetAl a ( A - ) gen) jXO 17.01 3*946 3.749 300 2* 14.42 3*767 3.746 ;320. 3° 16.35 3i321 3.755 '345 60.8 0.736 4° 16,28 2.929 3.755 400 80.4 0.981 5 14.36 2.871 3.747 475 94.8 1.154 6+ 11.98 2.840 3.739 566° 117.2 1.417 8 13.86 2.695 3.746 446 93.9 1.152 10 13,56 2.619 3.747 425 92.5 1.138 11+ 12.37 2.670 3.739 541 112.3 1.373 12+ 13.43 2.499 3.746 415 90.0 1.116 13+ 13.55 2.419 3.744 400 86.1 1.074 18 14.61 3.225 3.743 450 83.-6 1.001 19+ 9.25 3.073 3.732 610 132.5 1.572 20 13.45 2.881 3.747 490 105.0 1.273 22 15.15 2.914 3.744 417 85.6 1.042 24 14.83 3.L22 3.747 445 83.0 0.999 x present • < present o C present. Figure 5 Variation of the lattice parameter of the 6 phase . with carbon content. 17 3.760 L 3.450 U 3.740 3.730 k 3.720 FetAl Figure 6 Variation of the lattice parameter of the t phase with Fe:Al ratio. (For the carbon content of each alloy see table 1). 16 TABLE 3 Results of X-ray line intensity measurements. Specimen No.5 Specimen" No. 10 Line Index I (obs.) I (obs.) I (obs.) Ave. I calc. Diff. 1 13.0 11.8 12.4 10.4 19 .-• 2 • 3.0 2.7 2.9 2.9 0 100 100 100 100 ' -4 : ' 4 6 . 9 42.2 44.5 50.5 10 5 2.3 2.4 2.35 2.3 0.5 6 not measurable 0.5 -6 34.7 33.9 34.3 35.4 3.2 9 not measurable 1.3 -10 not measurable 0.3 -11 56.7 52.6 54.6 58.8 7 12 22.4 20.7 21.5 24.6 13 ,13 not measurable 1.1 Line -Index I (obs.) I.(obs.) I (obs.) Ave. I calc. % D i f f . 1 19.0 20.5 19.7 12.3 6o ; o 2 1.4 4.6 ; 3.0 2.8 6.7 3 100.0 < i o o ; o , , 100.0 100.0 -4 59.9 • 53.7 56.8 5 2 .3 7.9 5 3.0 3.0 .3.0 ' 2.4 ' 20.0 6 not measurable 0.4 -8 39.7 '•' 41.8 40.7 35.0 16.3 9 not measurable 1.3 -10 not measurable 0.2 -11 j 58.8 58.8 58.8 . 58.5 0.5 12 L K 2 1 24.2 24.2 24.6 1.6 1 3 not measurable 1.1 _ The measured and calculated intensities are based on a value 100 for line 3« 19 The measured valuer of Curie temperature, (5^ , and saturation magnetisation at absolute zero, $,, aa well as the calculated Bohr magneton nuffltar for moat of th© alloys may be found in table 2j the variation of these quantities with the FeiAl ratio or atomic peroent oarbon is shown in figures 7 tP 3.0 and 13 to 14. Soma of the 6 -1? curves are also shown in figures 11 and 12 for comparative purposes* From figures 1, 9 and 13 it is apparent that a l l ferromagnetic properties are decreased with increasing oarbon content, and, in fact, the decrease in Bohr magneton number with increasing oarbon content appears to be linear in the single phase region and with FetAl constant at 2,9» The slope of the linear portion of this curve is 0.133 -VAt, peroent Q, From this i t is easily shewn (Appendix 2a) that, between 14 and 15 percent carbon, the change in magnetisation corresponds to an effeotive decrease of 8.5 Bohr magnetons par oarbon atom. In figure 14 the Bohr magneton number for FeiAl slightly less than 3 appears to be constant when FetAl is changed* This, however, is deceptive since the amount of carbon in the alloys of this region is not constant* In order to obtain a wore exact picture of the variation of Bohr magneton number with the FeiAl ratio the assumption is made that the change with Qarbon is constant for a l l values of FeiAl* Since the change in Bohr n&fftjlan number with carbon content is Known at FetAl approximately 2.9 it is easy to correct the Bohr magneton number of each alloy to that expected for W &U°y with the same FeiAX ratio and 14*61 atomic peroent of carbon* The corrected relationship may be found in figure 150 Although the accuracy is decreased by the assumption made, a quantitative result may be obtained, as 20. 600 -Figure 7 Graph of Curie temperature, G e , versus atomic percent carbon for alloys of Fe-Al-C. (The FetAl ratio for the alloys shown is approximately 2.9). 21. 22. Figure 9 Graph of saturation magnetization, Go , versus atomic % carbon for alloys of Fe-Al-C with Fe:Al ratio approximately 2 . 9 . Fe:Al Figure 10 Variation of saturation magnetization, $« , with Fe:Al ratio for alloys of Fe-Al-C. (See table 1 for the amount of carbon in each alloy). 24, 100 80 70 60 50 40 30 20 10 Alloy Fe:Al % c 20 2.88 13.4 5 2.87 14i4 22 2.91 15.1 4 2.93 16.3 T5o~ "400" lob" 200 300 Temp. (°K) Figure 11 Variation of the- spontaneous magnetization, ff, with temperature for some alloys with different carbon content and Fe:Al approximately 2.9. 100 L 200 300 Temp. (°K) Figure 12 Variation of the spontaneous magnetization, § , with temperature for alloys of Fe-Al-C.having different Fe:Al ratios but approximately equal carbon content. Figure 13 Variation of Bohr magneton number with atomic percent carbon for alloys of Fe-Al-C. (Alloys 19, 6, 20, 5, 22, 4 have FetAl ratio 2.9 * 0.1, the remainder have various values of Fe:Al given in table 2). Fe:Al Figure 14 Variation of the Bohr magneton number with Fe:Al ratio for alloys of Fe-Al-C. (For carbon content of each alloy see table l ) . 28. Figure 15 Graph of Bohr magneton number versus Fe:Al ratio. All points have been corrected to the expected value for alloys with the same Fe:Al ratio and 14.61 atomic percent carbon. (The dotted lines are the expected curves for the cube corner iron atoms having Bohr magneton number 8 and 9 directed opposite to <P ) . 29. 1. Theoretical quantum curve j • 1/2 2. Theoretical quanturn curve j = 00 3. Measured curve for Mn3AlC 4. Measured curve for Fe-Al-C alloy No. 5. Figure 16 Magnetic moment per unit mass livided by saturation moment, % 0 » against absolute temperature divided by ferromagnetic Curie temperature Vet . (Ferro-magnetic Curie temperature for alloy No.5 was obtained by extrapolation of ff* versus T to 6»o ). 30. described in Appendix 2b. It is found that one atom of iron added when Fe:Al is less than 3 increases the magnetization by an amount equivalent to 5.4 * .5 /*a« For the ratio Fe:Al greater than 3, however, increasing iron decreases the magnetization, but the amount it does so is not yet known since the data in this region is very limited. v In the paramagnetic runs the balance deflection began increasing rapidly at 530°C, indicating that the ferromagnetic y phase was probably precipitating from the unstable £ . At temperatures above the Curie point of the y phase, the curve of 1/deflection against T has a slope much less than that required to extrapolate to the Curie point of the g phase. This may indicate that the £ phase has a curve of L/% against T which is concave toward the temperature axis, but this is not certain and paramagnetic measurements were on the whole unsuccessful. 31. V. DISCUSSION OF RESULTS AND CONCLUSIONS The proposed section of the phase diagram at 1050°C as deter-mined by metallographic techniques is much the same as that proposed by Morral at 1000°C. Since some of the alloys expected to be in the £ phase region were found to contain a small amount of y phase when quenched from 950°C, the change in the stability of the £ phase with temperature must be very rapid and there is no reason to doubt the validity of Morral*s diagram for alloys homogenized and quenched from 1000°C„ The x-ray line intensity measurements indicate that a very high degree of order does exist in the alloys even after quenching from 1050°C. The alloys have a face-centred cubic structure with the iron atoms taking up the face centres, aluminum the cube corners and carbon the body-centre positions 8 This is the result expected since i t agrees with the structure found for the analogous manganese alloys shown in figure 1. The increase in parameter with increasing carbon indicates perhaps that the lattice is strained slightly by its presence and may explain the limit of solubility at 15.7 atomic percent rather than the optimum of 20 atomic percent carbon,. The fact that the parameter does increase with carbon indicates that the aluminum and iron atoms are not as closely packed as possible and the parameter should not be expected to change with the Fe:Al ratio unless one of the atoms has a much greater size factor. Therefore, even though the change in the parameter with the FesAl ratio is not great, the iron and aluminum do not necessarily have equal size factors. If they were equal the diameter of the carbon atom at the cube centre would be 1.1 A°. This is very small since carbon usually has a much greater size a 32. factor and in diamond where a covalent bonding exists the interatomic distance is 1.54 A0. In the iron-aluminum-carbon alloys the carbon diameter is definitely larger than 1.1 A°; but possibly not greater than that found for diamond. This, as well as the persistence of structural order with temperature, leads to the conclusion that a bonding probably does exist between the carbon and iron. The most interesting feature of the magnetic properties of the 6 phase is the large variation in the magnetization with carbon content. The cause of this is not apparent but the following suggestion may be made. If there is a bonding between the carbon atom and its nearest iron neighbours, characteristics of the iron d states may be carried into the carbon atom. This would increase the opposition of equivalent character-istics of the iron states and perhaps neutralize the magnetic moment of electrons in the iron atoms. On the basis of arguments to be proposed the decrease in the magnetization due to this effect should correspond to a decrease of 6 Bohr magnetons per carbon atom. See Page 33. It was found experimentally,however, to be 8.5 at the FetAl ratio of 2.9. If the model proposed above is correct an alloy with the optimum composition of 20$ carbon should be expected to have no moment. Extrapolation of the curve to this value gives, however, a Bohr magneton number of 0.4. The magnetic behaviour of the alloys with changing aluminum con-tent is not yet certain but from the data available i t has been possible to postulate the following model. In an alloy with FetAl less than or equal to 3 and about 14.2 atomic percent of carbon there are effectively 3 types of iron atoms with their differences attributed to the number of nearest 33 carbon neighbours. Approximately #1/3 of the body centre positions contain carbon and by simple stat i s t i c s i t may be argued that 4/9 of the body centre positions have two nearest carbon neighbours, 4/9 have one nearest carbon and 1/9 have no nearest carbon. If the three types of iron atoms are given 1, 2, and 3 Bohr magnetons respectively the weighted mean Bohr magneton number is 1.67. Experimentally i t i s found to be 1.73. Each iron atom going into the l a t t i c e may have four valence electrons i n the 3 d states with their electron spins unbalanced. Assuming one of these electrons forms the bond with each nearest carbon atom and on the average one electron per iron atom i s paired with the valence electrons of the aluminum atom on the cube corner, there i s a basis for giving to each of the three types of iron atoms the Bohr magneton number which each has i n the proposed model. This model may be used to provide an explanation for the change i n Bohr magneton number with Fe:Al when this quantity i s less than 3. In this case the aluminum atoms i n excess of the optimum composition probably take up positions on the face centres where both adjacent cube centres are vacant. This i s another result of the iron-carbon bonding and i s i n agree-ment with the analogous manganese alloys. An iron atom replacing one of aluminum would necessarily replace one situated on a face centre to become an iron atom of the third type. It i s expected then that the increase i n magnetization should correspond to 3 Bohr magnetons per iron atom but there i s also another effect arising from the fact that an aluminum atom was removed i n the process. This aluminum has 3 valence electrons which may have the 34. a b i l i t y of pairing with 3 electrons i n the iron d states. The t o t a l increase i n magnetization should then correspond to 6 Bohr magnetons per iron atom. The experimentally measured value i s 5.4 * 0.5. When the Fe:Al ratio i s greater than 3 the excess atoms are iron and they must take up the cube corner positions in place of the aluminum. The decrease i n Bohr magneton number, as i n the isomorphous. manganese alloy, may be explained by the iron atoms on the cube corners having their magnetic moments directed opposite to the overall magnetization. The expected relationship between the Bohr magneton number and the Fe:Al ratio i s shown in dotted lines for the cube corner atoms having 8 and 9 Bohr magnetons directed opposite to the overall magnetization. It i s obvious that the experimental data i n this region i s insufficient but i f i t may be trusted at a l l the decrease i n magnetization with increasing iron must be due to another effect as well since iron i s not expected to have a Bohr magneton number greater than 4. It i s unfortunate that the shortage of time and many d i f f i c u l t i e s have not allowed for better results to be obtained i n this region. The graph of %•„ versus shown in. figure 16 for iron-aluminum-carbon alloy number 5 has somewhat the same shape as the theoretical, quantum curve for j s», whereas the corresponding curve for Mn^AlC has a quite different shape and i s close to the theoretical curve for j = '/j». This may mean that the magnetization i n the two alloys i s basically different although they are expected to have similar behaviour. The difference of course may easily be attributed to the lower carbon con-tent i n the iron alloy since this seems to greatly complicate the magnetic behaviour. In conclusion the iron-aluminum-carbon £ phase does have some of the peculiarities of the corresponding manganese-aluminum-carbon and manganese-zinc-carbon phases but experimental d i f f i c u l t i e s have hindered conclusive results. It i s very l i k e l y , however, that further work on the alloy system would be rewarding. 36 VI. APPENDICES APPENDIX 1. Particular D i f f i c u l t i e s Encountered i n Preparation of the Alloys. The Fe-Al-C system i s hard to investigate because the forma-tion of an oxide slag prevents the exact knowledge of the amount of aluminum alloyed with iron and carbon (Morral, 1934)o The different density of iron and aluminum also favours separation into layers of different composition which makes i t d i f f i c u l t to prepare homogeneous alloys. The f i r s t of these d i f f i c u l t i e s was practically eliminated in the present work by the use of a purified argon atmosphere around the melt. The alloys prepared did not appear to have any slag present. However, twelve alloys, which were prepared and assayed, were found to be very far off composition and the metallographic observations did not agree with other results. The cause of this was thought to be a contaminated argon supply and many of the alloys were discarded for this reason. The homogeneity of the alloys was checked by measuring the spontaneous magnetization at room temperature for samples taken from different parts of the melt. Several of the alloys were found to have their magnetizations varying by as much as 10 percent and were therefore discarded as well. It was found that alloys with FejAl, greater than 3, were particularly hard to prepare although they do not obviously have any physical properties which di f f e r from those of alloys with FetAl less than 3. 37. APPENDIX 2. a. Calculation of the change in Bohr magneton number with changing carbon content. In the single phase e region with FetAl equal to 2.9 * 0.1 the graph of Bohr magneton number versus atomic percent carbon is a straight line with slope 0.133 S*a /At. percent C. At 14 atomic percent carbon an alloy with FetAl equal to 2.9 has 63.95 percent of its atoms iron and 22.05 percent aluminum. The Bohr magneton number of this alloy as obtained from the graph would be 1.203, and 100 atoms of alloy, 14 of which were carbon, would have a total Bohr magneton number of 63.95 x 1.203 • 76.92 /<* . In the same way, at 15 percent carbon 63.19 percent of the atoms are iron and 100 atoms of alloy would have a total Bohr magneton number of 63.19 x (1.203 0.133) " 67.61 MB and in this case 15 of the atoms would be carbon. The difference i n Bohr magneton number between the two alloys is 9.31 /** and this must be attributed to the one atom of carbon difference and also to the slight decrease in iron content of 0.75 atoms with an average of 1.13 > u „ per iron atom. The result then is that there is an effective decrease of 8,5 * 0.5 Bohr magnetons per carbon atom. The uncertainty i n this value is made larger than the measurements indicate to be necessary in view of the fact that even in very simple systems the determination of Bohr magneton number cannot be done accurately. b. Calculation of the change in Bohr magneton number with changing FetAl ratio. The graph in figure 15 showing the relationship between FetAl and Bohr magneton number indicates that the latter changes from 1.00 fiK 38. at Fe:Al = 2.58 to 1.17 at Fe:Al • 3 and 14.61 atomic percent of carbon have a total Bohr magneton number of 64.04 x 1.17 • 75.0 /<8> whereas 100 atoms of an alloy with Fe:Al = 2.58 and the same carbon content have a Bohr magneton number of 61.70 x 1.00 = 61.7 /<» . The difference 13.3 must be attributed to the increase of 2.34 iron atoms. That is,the magnetization increases an amount which corresponds apparently to 5.4 * 0.5 Bohr magnetons for every iron atom added when Fe:Al is less than 3. 39 VII. BIBLIOGRAPHY 1. R. G. Butters and H. P. Myers, Phil. Mag., Ser. 7, Jj6, p. 895, 1955 (A). 2. R. G. Butters and H. P. Myers, Phil. Mag., Ser. 7, 46, p. 132, 1955 (B). 3. R. W. James, "The Optical Principles of the Diffraction of X-rays," London: G. Bell and Sons, (1948). 4* K. Hoselitz, "Ferromagnetic Properties of Metals and Alloys," pp. 142-143, (1952). 5. W. Hume-Rothery, "The Structure and Property of Metals and Alloys," p. 119, (1945). 6. 0. von Keil and 0. Jungwirth, "Archiv fur das Eisenhutenwesen," 5_, p. 221, (1931). 7. G. Kidson, Master1s Thesis, University of British Columbia, p. 30, (1953). 8. F. R. Morral, The Journal of the Iron and Steel Institute, Vol. CXXX, pp. 419-427. 9. L. Neel, Ann. Phys., 1948 , 3., p. 137 10. J. L. Snoek, "Research in Holland," p. 99, (1945). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0106335/manifest

Comment

Related Items