L. H. REYERSON, J. E. WERTZ,W. WELTNER,JR.,AND H. WHITEHURST
1334
Vol. 61
SORPTION AND MAGNETIC SUSCEPTIBILITY STUDIES ON A GRAPHITE-BROMINE SYSTEM' BY L. H. REYERSON, JOHN E. WERTZ,W. WELTNER,JR.,AND HARRY WHITEHURST Contribution from the School of Chemistry of the University of Minnesota, Minneapolis, Minnesota Received February Wb, 1067
Adsorption-desorption isotherms were obtained a t 28.8' for the system bromine-graphite. Magnetic susceptibilities for this system were measured a t each oint on the isotherms. At 28.8' about 500 mg. of bromine was adsorbed per gram at saturation. At the same time the Jarnagnetism of the graphite-bromine system had fallen to about 6% of its original value, Obviously the sorption of bromine had markedly altered the structure of the graphite. Furthermore at zero pressure of bromine the graphite retained almost one hundred mg. of the vapor per gram. Since the diamagnetism did not return to its original value at this point i t seema very probable that the bromine molecules were retained in the swollen graphite structure.
While studying the magnetic susceptibility of solids during the process of sorbing such gases as the oxides of nitrogen2 it was noted that the susceptibility of graphite, during the sorption of nitrogen dioxide, was altered considerably more than predicted from the sum of the susceptibilities of the two substances. This suggested a disturbance of the graphite structure. Some swelling of the crystals of graphite in the presence of the nitrogen dioxide seemed to confirm the above suggestion. Earlier a number of investigators had demonstrated that graphite swelled in the presence of such substances as liquid potassium, sulfuric acid, bromine and antimony trichloride. Allied t o these are such systems as graphite-ferric chloride, graphite-fluorine and graphite-oxygen. Rudorff3 showed that bromine was able to enter alternate spaces between planes in graphite in a quantity such that the ratio of carbon t o bromine atoms was 8:1, increasing the normal spacing to 7.05 A. while the adjacent spacings were 3.35 A. In view of this work, together with the difficulty of measuring the degree of diamerization of nitrogen dioxide on graphite, it was decided to determine a complete isotherm for the sorption of bromine vapor by graphite. At the same time the susceptibility of the bromine-graphite sample at each equilibrium point would be measured. It was hoped that the results so obtained might serve to settle, at least in part, some of the questions raised with regard t o the behavior of graphite toward a number of substances. Experimental A sample of the same graphite, as used in the previous investigation,* was placed in a thin walled spherical Pyrex bulb suspended from a quartz-spiral spring balance. The whole glass system, containing this spiral, was mounted in such a manner, on the frame supporting the magnet, that the precision lift was able to raise or lower the magnet until the bulb containing the graphite was located a t the point of maximum magnetic flux. As the graphite took up or gave off bromine the extension of the spring permitted the calculation of the amount sorbed a t each pressure of bromine. Following the determination of an equilibrium value of the amount sorbed a t a given pressure the magnetic field was turned on and the magnet moved to the point of maximum effect on the sample. The additional change in the extension of the quartz spring made i t possible to calculate the change in susceptibility due to the change in the amount of bromine sorbed by the graphite. (1) This investigation was supported in part by the Office of Naval Re 8earch (2) L. H. Reyeraon and John E. Wertz, THISJOURNAL, 63. 232
.
(1949).
(3) W. Rfidorff, 2. anoru. aZ2gern. Chern., 245, 383 (19411.
A constant vapor pressure of bromine was maintained in the system, while a given equilibrium point on the isotherm was being established, by keeping the reservoir holding the liquid bromine at a constant temperature in a cryostat similar to that of Scott and Brickwedde.' Carefully purified bromine was first distilled over PlO6 and then vacuum distilled into the reservoir which was separated from the rest of the system by a glass septum until the graphite had been thoroughly outgassed and the system made ready for the sorption process to begin. The section of the glass system containing the spiral and the sample was water jacketed so that by controlling the temperature of the circulating water a constant temperature was maintained for the spring balance and sample. The sorption isotherms here reported were carried out a t 28.8'. Equilibrium was established rather rapidly for the first four or five oints on the initial sorption isotherm. From there on it ogen took more than 24 hours of time for the determination of a single equilibrium point. The completed sorption-desorption isotherms as given here required more than one hundred days with numerous measurements being made during each 24 hour period. Sorption values were determined up to about the saturation pressure of bromine. This was followed by a desorption isotherm ending a t zero bromine vapor pressure. At this equilibrium point almost one hundred milligrams of bromine remained sorbed per gram of graphite. From this point two additional sorption-desorption isotherms were determined, ending a t the saturated vapor pressure of bromine. Magnetic susceptibility values were obtained at each of the sorption points.
Results The experimental data obtained during this long run are given in Fig. 1. The curves in the lower section of the figure show the sorption-desorption isotherms while the upper set of curves plot the percentage of the diamagnetism of the graphite found for each equilibrium sorption or desorption point. The open circles give the values for the first sorption isotherm, the first desorption isotherm, and the second sorption isotherm. They cover the first isotherm carried to about the saturated vapor pressure of bromine together with the first complete desorption-sorption cycle which followed. The half-filled circles give the results for the second complete desorption-sorption cycle and the dark circles cover the third complete cycle ending a t the saturation pressure of the bromine. Curve I represents the first sorption uptake while curves I1 and I11 represent the subsequent desorptionsorption cycles, respectively. Several more points were actually obtained for curve I below 25 mm. vapor pressure of bromine but they did not change the shape of this part of the isotherm so were left off in order not to crowd the points on the curve. This section of the iso(4) R. B. Scott and F. G. Brickwedde, J . Research Null. BUT Standards, 6 , 401 (1931).
Oct., 1957
SORPTION AND MAGNETIC SUSCEPTIBILITY OF GRAPHITE-BROMINE
therm resembles a Langmuir type of adsorption because the curve flattens out and appeared to reach a monolayer coverage of the graphite crystals. The magnetic susceptibility measurements tend to confirm this assumption because until a vapor pressure of between 10 and 20 mm. was reached the susceptibility values remained constant within the limits of reproducibility of the magnetic field. Furthermore, by taking the cross-sectional area of bromineofromliquid close packing to be twenty-one square Angstroms and assuming monolayer coverage one obtains a surface area for the graphite which agrees very well with 3.7 m.2/g. area obtained by nitrogen adsorption at liquid nitrogen temperatures, When the vapor pressure of bromine had increased to about 25 m.m.p. a marked change in the character of the isotherm took place. The changes in susceptibility and the swelling of the graphite crystals indicated that bromine molecules began entering the layered structure. Obviously an irreversible alteration of the graphite structure began to take place following the initial monolayer coverage. The residual amount of bromine retained a t the end of each desorption run and the limiting value of the susceptibility at low bromine vapor pressure both confirm this. Additional evidence for such a permanent change was found when it was observed that excellent graphite crystals increased 60% in thickness during sorption and then diminished only half of this amount upon reducing the vapor pressure of bromine to zero. Curve I is entirely distinct from curves I1 and 111. However, curve 111 which is a composite of three sorptions each of which followed a desorption does parallel somewhat the upper part of curve I. Once the initial isotherm was traversed, the addition and removal of bromine proceeded in a reproducible fashion. It is remarkable that' such definite hysteresis should persist down to essentially zero pressure. The isotherms found in this study differ considerably from those determined by Juza and his coworkers5 at 0". They found that their graphite continued to sorb bromine strongly as saturation pressures were approached. In the present work little further sorption was found above eighty per cent. of the saturation pressure of bromine. Curve I appears to be a type IV isotherm as considered in connection with BET theory. Juza reported that the graphite he used sorbed 0.6 g. of bromine per gram a t saturation while Goldsmitha gave a figure of 0.7 g. per gram. The limit found in this study was almost exactly 0.5 g. of bromine per gram of graphite. About 17% of that sorbed at the saturated pressure remained sorbed at zero pressure of bromine. This indicated that at least two types of sorption are found a t 28.8'. Both types altered the susceptibility of the graphite. (5)
R. Juse, H. Lubbe and L. Heinlein, Z. anoro. Cham., 268, 105
(1949). (6)
M.Goldsmith, J . Chem. P h y s . , 18,523 (1950).
1335
At saturated vapor pressure of bromine, the susceptibility had fallen to about 6% of its original value. As the vapor pressure of bromine was re-
b l
0
v
I
I
I
I
v
100 IM 200 BROMINE PRESSURE IN MILLIMETERS.
50
Fig. 1.-Adsorption
KO
of bromine on graphite.
duced very little change in the susceptibility or in the amount sorbed was noted until the vapor pressure had fallen to about 50 mm. From this point on the susceptibility increased and the amount sorbed decreased. At the zero pressure equilibrium point the susceptibility remained at 66% of the original value. One concludes from these data that bromine enters the lattice of graphite as soon as the crystalline surface has sorbed a monolayer of gas. From this point on the susceptibility falls as the bromine enters the lattice until it is only 6% of the original value. When the gas pressure is again reduced to zero, 17% of the total sorbed bromine remains in the graphite in what appears to be a chemically bonded condition. It is likely that the interstitial bromine interacts with the conduction electrons to greatly reduce the diamagnetic susceptibility of the system. DISCUSSION J. J. BIKERMAN.-I was glad to hear the emphasis put by Professor Reyerson on the alteration of the adsorbent by the adsorption, which alteration is too often disregarded. Did the authors try to apply any of the theories of the magnetic susceptibility of mixtures to their results? With the aid of these theories it might be possible to determine in what form bromine is present or how it is distributed in the system. L. H. REYERSON.-NOt as yet,