variety of uses other than as a chemical reagent. It is
may be impregnated with molten sulfur (5). Attempts have been made to use it to consolidate quicksand, etc. (I). Molten sul
fur with (4), and without (12), viscosity-modifying substances, has been proposed for use as a drilling fluid to replace drilling mud and to cement oil-well casings in the petroleum industry, and also to confine and make fast various types of earth forma
tions (IS). These uses undoubtedly could be augmented if the viscosity of sulfur at elevated temperatures could be modified.
T h e m a n ifo ld u ses o f s u lfu r u n d o u b t e d ly w o u ld b e a u g
The viscosity of pure sulfur (Figure 1) increases sharply with rise in temperature above 160° C., and reaches a maximum of
■93,200 centipoises at 186-188°. As the temperature rises still more, the viscosity falls to about 2600 centipoises at 300 ° C. and to about 100 centipoises at the boiling point, 444.6°. These high viscosities often make difficult the handling of molten sulfur.
They bar the use of sulfur as a heat transfer medium in an im
portant temperature range. They interfere with the use of ele
mental sulfur in chemical reactions at high temperatures. They retard penetration.or impregnation operations.
However, sulfur also has many good qualities. It is cheap and high in purity in all commercial forms. It has a low vapor pressure almost throughout its liquid range. It has interesting
thermal properties. Be known which modify the viscosity of sulfur. The most effective are the halogens, hydrogen sulfide, hydrogen persulfides, and the hydrocarbons. The latter or their derivatives function through the presence of hydrogen sulfide and hydrogen persulfides result
ing from their reaction with sulfur.
V I S C O S I T Y D E T E R M IN A T IO N
The viscosities were determined in open and closed systems.
In the open system (S) volatile matter was allowed to escape.
In the closed system special viscometers of the type shown in Figure 2 were used. After charging, the instrument was placed in an air bath (3) and mounted in a cradle which could be rotated through 360° C. The temperature was shown by an iron-con- stantan thermocouple fastened to the narrow end of the viscom
eter. The viscometer was rotated from the filled position through 180° C., and the time of emptying the larger bulb noted. The viscometer was calibrated by a water-glycerol mixture of known viscosity. The weight of sulfur in the mixtures was about 35 grams, and the free air space in the sealed viscometer was approximately 50 cc.
The materials were sufficiently pure for these determinations. The sulfur was purified according to a method previously described (2). Chlorine was added in the form of sulfur chloride which was analyzed for its chlorine con
tent. Bromine and iodine were added in elemental form. All mixtures are in terms of weight per cent, and their densities were assumed to be the same as pure sulfur.
In the closed system experiments, hy
drogen sulfide was added to the sulfur as liquid and in the form of hydrogen persulfide. In the former case the vis
cometer and sulfur were chilled by dry ice, and a measured volume of liquid be analyzed easily and accurately for its hydrogen sulfide content (14). The con
centrations given indicate the weight F i g u r e 2 . V i s
c o m e te r U sed fo r th e C losed S y stem 39
40 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y Vol. 38, No. 1
SULFUR-CHLORINE MIXTURES CLOSED SYSTEM
— X — > ^ 1 . 5 %
---2 . 0 £
180 2 0 0 2 2 0 2 4 0 2 6 0 2 8 0 3 0 0 3 2 0 3 4 0 3 6 0 3 8 0 4 0 0 TEMPERATURE «C.
F ig u re 3. V iscosity o f S u lfu r C o n ta in in g H alogen s
ratio of hydrogen sulfido to the total weight of the mixture in the closed system. Most of the hydrogen sulfide remains in the gase
ous state and only a fraction of it dissolves in the liquid sulfur.
The hydrogen persulfides dissolve quickly in sulfur at about 120° C. with no decomposition; but with increasing tempera
ture, decomposition sets in and liberates hydrogen sulfide.
E F F E C T O F H A L O G E N S
Figure 3 presents the variation in viscosity with temperature of sulfur containing various concentrations of halogens. Com
parison of these curves with Figure 1 brings out the enormous modifying effect of these elements on the viscosity of sulfur. Of the three elements, chlorine is the most effective. Mixtures containing 2% chlorine give values mainly under 15 centipoises throughout the entire temperature range. Increasing concen
trations give lower viscosities which tend to bo independent of temperature, as shown by the 3, 4, and 5% curves. These curves show that the viscosity of pure sulfur can be reduced easily from its maximum of 93,200 centipoises to about 10 centi
poises. They also indicate that, with each increment in concen
tration, the drop in viscosity becomes smaller and tends toward a limiting value. This holds for all known substances which modify the viscosity of sulfur. Although iodine is the least effective of the halogens, it substantially reduces the viscosity.
Bromine takes an intermediate position, being considerably more effective than iodine but proportionately less effective than chlorine.
As is to be expected for equal concentrations, values given for closed systems are lower than for open systems. The difference, 2!
Hydrogen sulfide (Figure 5) is at least as effective as chlorine, if not more so, in reducing the viscosity of sul
fur; but because of its rela
tively low solubility in sul
fur, high hydrogen sulfide pressures are necessary to bring about very low viscosi
ties. Curve 1 shows the vis
cosity values obtained by bub
bling hydrogen sulfide constantly through sulfur in an open system. The remaining solid curves are for closed systems in which the hydrogen sulfide was added in the form of hydrogen persulfide. The broken curves were given by a closed system in which the hydrogen sulfide was added as a liquid.
E F F E C T O F O R G A N IC M A T E R IA L
A number of hydrocarbons and their derivatives have been proposed for modifying the viscosity of sulfur above 100° C.
(4, 5, 10). Their effectiveness has been shown (S) to be due to
Wul 24C
<0
£ 160 80 0
TIME IN-HOURS
F ig u re 4. V iscosity C h an ges a t 200° C . w ith T im e 0 .7 5 % c h lo r in e 3. 0 .2 5 % io d in e 5. 1.0 % io d in e 1.8 % c h lo r in e 4. 0.5 % io d in e 6. 0 .5 % b r o m in e
however, is not so great as was anticipated. The loss of halogens, free and combined, by volatilization from an open system is quite slow; Figure 4 shows the change in viscosity o f sulfur-halogen mixtures with time. The mixtures were maintained at 190- 200° C. for the time indicated, and the viscosity determina
tions were made at 200° C.
The curves clearly indicate that, while chlorine is most effective in reducing the vis
cosity, its volatility in the form o f sulfur ch loride is greatest. With respect to per
sistency, iodine is much better than chlorine and bromine is better than either. These re
sults are in keeping with the following boiling points: sul
fur ch loride, 138°; iodine, 184°; sulfur bromide, 58° C.
at 0.22 mm.
E F F E C T O F H Y D R O G E N S U L F ID E A N D P E R S U L F ID E S SULFUR-TODINE MIXTURES
CLOSED SYSTEM OPEN SYSTEM
SULFUR-BROMINE MIXTURES CLOSED SYSTEM OPEN SYSTEM toUI
V) 8 0 5
Q-January, 1946 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y 41
T E M P E R A T U R E °C.
Figure 5. Effect o f H ydrogen Sulfide on Viscosity o f Sulfur
the presence of hydrogen sulfide and hydrogen persulfides result
ing from the action of sulfur on the hydrocarbons.
Sulfur, as mined by the Frasch process, is always associated with relatively minute but varying amounts of oil. Hence, its commercial forms are never free of oil. Simply heating this sul
fur above 180° C. quickly produces sufficient hydrogen sulfide and hydrogen persulfides to reduce viscosity greatly. This di
rect and easy procedure m ay be made to give sufficiently low vis
cosities for some purposes.
As an example of this form of heat treat
ment, the following experiment is cited: A sample of commercial sulfur containing 0.038%
oil was heated in an open system from 125° to 260° C. in 1.5 hours and kept at this tempera
ture for an additional half hour. The viscosities 360 given by the resulting material on cooling to
170° C. are shown by curve 1, Figure 6. Com parison of this curve with Figure 1 shows how effective this treatment can be. Redetermined values with rising temperature are given by curve 2. The mixture was again cooled to and kept at 160° C. for 14 hours, and viscosities were determined with rising temperatures (curve 3). This cooling and reheating process was twice repeated, and the respective data are shown by curves 4 and 5. The relative positions of the curves show that prolonged heating in an open system results in a slow loss of hydrogen sul
fide and persulfides with resultant increase in viscosities. In a closed system low viscosities can be maintained indefinitely.
Viscosities for temperatures above the boiling point of sulfur may be roughly estimated from Figures 7, 8, and 9 which were ww
w 90 5a.
p z
111 u
wUJ W O
CL
TEMPERATURE °C.
Figure 6. Effect o f Traces o f Oil on V iscosity o f Su lfu r 280
Power P la n t o f Texas G u lf Su lph u r C om p an y a t N cw gu lf, T exas, w ith V a ts o f M in e d S u lfu r a t U pper R ig h t
42 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y V ol. 38, No. 1
w
LU
Oa.
LU
Figure 7. V iscosity a t Elevated T em p eratu res o f Su lfu r C o n ta in
ing H ydrogen Sulfide
2 3 4 5 6 7 TEM PERATU RE X100°C.
Figure 8. Viscosity o f Su lfu r at Elevated T em p eratu res 1 . S u lfu r + H iS a t a t m o s p h e r i c p r e s
s u r e 2 . P u r e s u lfu r
Figure 9. V iscosity a t Elevated T em p eratu res o f S u lfu r C on tain in g
H alogens 1.
2.
2 % c h lo r in e 1 .5 % c h lo r in e
6 % io d in e 3 % i o d in e
1 .6 8 % II2S an II2S*
1.7 % l i q u id I liS 1.1 % l iq u id II*S 0 .9 5 % II3S an IU S X 0 .9 % l i q u id HaS
obtained by plotting the appropriate values given by the above systems on log-log paper. The effect of pressure on sulfur is to increase the viscosity (11 ). This tendency is reflected in the extrapolated values o f Figures 7 and 9 for closed systems as com
pared with those from Figure 8 based on values given by an open system.
C O R R O S IO N B Y H A L O G E N M I X T U R E S
These halogen mixtures would be expected to be highly corro
sive, but preliminary experiments with a number of metals in
dicate strong resistance to disintegration. The tests were carried
Ta b l e I. Co r r o s i o n Re s i s t a n c e o p Me t a l s t o Su l f u r a n d Su l f u r- Ha l o g e n Mi x t u r e s
T em p., 1 8 -8 steel
°C . A lum inum 1S-S steel + 3 M o
Pure sullur 210 N one 0 .0 0 00 0 2 N one
410 N one 0 .0 0 0 0 8 0 .1 3 % gain in w t.&
Sulfur + 1 %
iodine 210 0.0 0 00 0 3 0.0 0 00 1
410 0 .0 0 0 0 0 5 1 .1 % gain in wt.&
0 .1 % gain in w t.b Sulfur + 0 .7 %
brom ine 210 0.0 0 00 0 9 0 .0 0 00 0 4 0 .0 0 00 2 410 0.0 0 04 7 0.0 0 01 9 0.0 0 02 9 Sulfur + 1 .7 %
chlorine 210 0 .0 0 1 0.0 0 00 1 2
410 0.0 0 1 0 .0 0 0 4 0.0ÓÓÍ6
H astelloy C
0.0 0 00 0 7 0.0 0 00 5
0 TJ. S. Steel R esearch Lab. specifications: fully resistant, penetration less than 0.00035 inch per m on th ; satisfactorily resistant, 0 .0 0 0 3 5 -0 .0 0 3 5 in ch ; fairly resistant, 0 .0 0 3 5 -0 .0 1 0 inch.
& M eta l sound, n o swelling.
out in sealed glass tubes. Test pieces were completely immersed in the mixture at 210° and 410° C. for 30 to 40 days. Some o f the results are shown in Table I.
The sulfur-chlorine mixtures are particularly destructive- toward aluminum. A t the end of the experiments the aluminum was coated with a thick, adherent, protective film. This film was a mixed aluminum chloride and sulfide. A t 410° C. the sulfur-bromine mixture deposited the same type o f coating on the- aluminum. The iodine-sulfur mixtures appear to be the least corrosive. The easy hydrolysis of the sulfur halides makes it imperative that water in any form be rigorously excluded.
These values are given merely to indicate that a number o f common metals under these conditions satisfactorily resist the action of these sulfur-halogen mixtures. So many factors enter the corrosion picture that one cannot predict with certainty the behavior of a metal under another set o f conditions. Recent ad
vances in the fabrication o f highly resistant metals to heat, pressure, and corrosion indicate that all adverse conditions can be met successfully.
D IS C U S S IO N
Substances which reduce the viscosity o f sulfur above 160° C . do so through chemical reaction with the sulfur. The simi
larity in the shape o f the curves for the various substances indicates- the same underlying mechanism. In the temperature range below 160° C. the molecular complexity and structure o f the liquid is relatively simple, consisting mainly of eight-membered puckered rings. In this region the viscosity ranges from about 12 centipoises at 120° C. through a minimum o f about 7 centi- poises at 157° and back again to 12 centipoises at 159°, and these values remain essentially unchanged by the addition o f viscosity-m odifying substances. A bove 160° C. these rings rup
ture to form long chains which increase in length up to 187°.
January, 1946 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y 43
This polymerization of sulfur into long chains brings about the increase in viscosity with rising temperature (3, 7). Increase in temperature beyond 187° results in a shortening of the chains and, consequently, a falling off in the viscosity as shown in Figure 1.
Reduction in the viscosity by the halogens, hydrogen sulfide, and hydrogen persulfides must be duo to a reaction which shortens the chains. This scission is believed to take place with the halo
gen atoms taking terminal positions of the segments. Thus for chlorine we would have:
Cl—S—S— S • • • • S—S — C l or C l—S—S— S • • ■ S— S —S In the same manner hydrogen sulfide and persulfides shorten the chains with hydrogen as the terminal atoms of the segments.
Such a mechanism would also explain why these substances per
sist so tenaciously in the liquid sulfur even when the mixture is kept far above their boiling points.
There appears to be a great deal of uncertainty as to the direct combination of sulfur with iodine (9 ). While sulfur iodides may be made by indirect methods, the direct addition of iodine to above mechanism for the reduction of viscosity is sound.
L IT E R A T U R E C IT E D
(9) Mellor, J. W., "Comprehensive Treatise on Inorganic and Theo
retical Chemistry” , Vol. X, p. 653 (1930).
(10) Potty, G. M., Univ. of Pittsburgh, Bull.30, 2 (Nov. 15, 1935). bcnzoplicnone arc described as new exam ples o f th e a ppli
cation o f su ch cata lysts to con ven tion al reaction s o f the Friedel-C rafts type. •
I
^ R IE D E L -C rn fts chemistry was founded in 1877 when Friedel . and Crafts discovered the condensation of aromatic hydrocarbons and alkyl or acyl halides with aluminum chloride. Since then, halides of aluminum, tin, zinc, iron, and other metals, and acids such as sulfuric, phosphoric, and anhydrous hydrofluoric have been found to catalyze a wide variety o f condensation reac
tions. The scope of the condensation reactions has been greatly expanded in the last twenty-five years. The alkylation o f paraf
fins and naphthenes, particularly of isoparaffins, is a recent de
velopment commercialized on a large scale in the production of high-octane hydrocarbons and fuels. These same catalysts have been applied to other types of reactions such as isomerization, transfer of radicals, and cracking.
The application o f silica-alumina and certain homogeneous cat
alysts to reactions of these types presents new fields of scientific endeavor.
S IL IC A -A L U M IN A C A T A L Y S T S
Active heterogeneous catalysts containing silica and alumina are produced either by activation o f some natural clays or by synthesis. Silica gel, notwithstanding its enormous surface, does not catalyze the reactions described in the present paper. A small proportion o f alumina, of the order of 1% by weight o f the silica,
is sufficient to produce an active catalyst. Commercial cracking catalysts contain approximately 10% alumina. Oxides such as thoria or zirconia can be substituted for the alumina. Silica- alumina catalysts were first developed for the catalytic cracking of petroleum oils. Approximately 1,000,000 barrels of oil are now cracked daily over these catalysts.
The alkylation o f aromatic hydrocarbons with olefins, long es
tablished in Friedel-Crafts and strong acid syntheses, was the first application o f silica-alumina to condensations o f the Friedel- Crafts typm Michel (11) described the condensation o f naphtha
lene with propylene under pressure over fuller’s earth to pro
duce tetraisopropylnaphthalene. Schollkopf (19) alkylated naphthalene with ethylene at 230° C. under 20-40 atmospheres pressure over an activated hydrosilicate catalyst. Sachanen and O’K elly (17) described the alkylation o f benzene with propylene, butylenes, and amylenes over silica-alumina at 450° C. and 100 atmospheres. Under these conditions the alkylation proceeded sm oothly but was accompanied by partial cracking of the paraf- finic side chains. As a result, toluene, ethylbcnzene, and xylenes were produced in substantial yields.
Destructive alkylation reactions catalyzed by aluminum chlo
ride, as observed by Ipatieff and co-workers (7), were carried out over silica-alumina catalysts by Sachanen and Davis (15). These investigators reacted benzene with pcntanes over an activated clay for 45 minutes at 480° C. and 1050 pounds per square inch.
Twenty-eight per cent (by weight of benzene charged) of alkyl- benzenes boiling from 105-210° C. was produced. The applica
tion o f silica-alumina to reactions formerly catalyzed by Friedel- Crafts and strong acid catalysts was increased in scope by Hans
ford, Myers, and Sachanen (4) and by Thomas, Hoekstra, and Pinkston (S 3 ). These investigators dealkylated alkylaromatic hydrocarbons in the presence o f silica-alumina at 450-550° C.
An example o f these reactions was the conversion of ethylbcnzene to benzene and ethylene.