Elemental Sulfur in Crude Oil B. H. ECCLESTON, X Q R I L Y N MORRISON, AND H. M. SMITH Petroleum Experiment Station, Bureau of Mines, Bartlesville, Okla. The question as to whether crude oil contains elemental or free sulfur has never been definitely settled, although there are many references in the literature to the presence of elemental sulfur in distillates and as condensed crystals in distillation equipment. In connection with studies on sulfur compounds in petroleum, the presence and behavior of elemental sulfur in crude petroleum are important. Elemental sulfur was determined in 17 crude oils by a polarographic method which is believed to be accurate and to respond correctly to the presence of elemental sulfur in the oil. Particular attention was paid to the sampling and to the handling of the samples after they w-ere received in the laboratory. The data obtained should prove very helpful in studies of the thermal reactions of sulfur-containing crude oils and explain, in part at least, the large evolution of hydrogen sulfide often observed at temperatures in the vicinity of 200OC. (392' F.) when crude oils are distilled.
T
H E literature contains many references which indicate that a sulfur-bearing crude oil is very likely to evolve hydrogen sulfide, sometimes in copious quantities, during distillation in the temperature range of 150" to 250' C. (302' t o 482" F.). Inasmuch as distillation is one of the basic tools in separating the components of petroleum, API Research Project 48-4, on the production, isolation, and purification of sulfur compounds and measurement of their properties wished to use distillation whenever it could be applied without causing essential change or decomposition in the sulfur-bearing components of the oil. I n the early work of the project, it was soon discovered that there was considerable difference in the types of sulfur compounds present in distillates from a given crude oil, depending upon whether the distillation had been conducted under atmospheric or reduced pressure conditions (8). This observation led to the development of a stability test ( I ) , n-hereby the crude oil is boiled under reflux conditions for a selected interval and any evolved gases, such as hydrogen sulfide and mercaptans (thiols), are passed through suitable absorbers. After a given time, a small amount of distillate is withdrawn, raising the refluxing temperature, and again the material is refluxed a t a new temperature level for a considerable period, and the gases evolved are measured. A study of several crude oils by this procedure showed that large amounts of hydrogen sulfide \\-ere evolved from certain crude oils a t about 200" C. (392" F.). Further elemental sulfur v a s found in some of the distillates, in distillations of sulfurbearing crude oils, and in a t least two instances sulfur crystals collected in fairly large quantities in the condenser of the apparatusk. ,From these observations it was inferred that in many, if not all, instancep, the hydrogen sulfide evolution a t around 200" C. (392' F.) was caused by reaction of elemental sulfur m-ith certain components of the oil. This theory emphasized the need for considering the presence of elemental sulfur in crude oil. Up to the present the analytical determination of elemental sulfur in crude oil has been beset with difficulties and uncertainties. It is for this reason, perhaps, that the presence of elemental sulfur in crude oils has been difficult to establish and that any announcement of its presence has met with skepticism ( 4 , 7 ) . Because hydrogen sulfide is often as-
sociated with crude oil and it readily reacts with oxygen to yield elemental sulfur, this skepticism is probably warranted. Mabery reported ( 1 O j finding sublimed sulfur in the neck of a condenser when distilling Ohio crude oil. Richardson and Wallace ( 1 7 )also found elemental sulfur in Beaumont, Tex., crude oil, but their results are clouded by the method of isolation, in one step of which the crude oil was blon-n tvith air. They concluded that the evolution of hydrogen sulfide from a hydrogen sulfidefree crude oil, when the oil is heated to 200' C. is caused by the presence of elemental sulfur in the oil, and in this the present authors are in agreement. Birch and Sorris ( 2 ) found elemental sulfur in the fractions from an Iranian crude oil when the distillation temperature reached about 120" C. (248" F.) but attributed this to decomposition of sulfur compounds. Other investigators (12, 19) have also reported finding elemental sulfur occurring naturally in crude oil. Hon-ever, beeawe of the conditions under which the determinations have been made or because of uncertainty regarding the history of the damples, no authentic data. supporting the presence of elemental sulfur in crude oil have been found. However, the available data, considered with those developed by API Research Project 48-4, n-ere interesting enough t o encourage investigation of crude oils for elemental sulfur by the polarographic method, which had been reported in the literature as successful on petroleum products. One of the first papers, if not the first, dexribing polarographic determination of a sulfur compound was published by Brdicka ( 3 ) in 1933; he successfully determined cystine, a disulfide of biological importance of the formula: HCO~CHXHZCH~-S--S-CH2NH2CHCOZH. The method of Brdicka was further investigated by Rosenthal (It?), n-ho applied it to determination of disu! fide and mercaptan groups in biological materials. Pro?!.! (13, 26) published two papers in 1940 on the polarographic dctr I mination of elemental sulfur and accelerators in rubber prc :ucts, and two further papers (14, 16) in 1946 and 1947. Also in 1947, Gerber (6)applied t,he method of Proske to petroleum products for determining hydrogen sulfide and mercaptans; this was followed 3 years later (6) by a discussion of the polarographic* determination of elemental sulfur and disulfides in gasoline.. . The first American publication found that discusses the polarographic determination of elemental sulfur in petroleum fractions was by Hall (9), and his basic procedure has been follo~~-ed in the present work. APPARATUS AND REAGENTS
Sargent Model X X I recording polarograph, H-type electroIysis cell, and mercury-pool cell. Constant temperature bath held a t 25" C. Capillary having the following characteristics: drop time 3.97 seconds (in distilled water), rate of mercury flow 2.72 mg. per second, and mercury column height 44.9 cm. Pyridine, Mallinckrodt analytical reagent. Methanol, absolute. Sulfur, U.S.P. In polarographic determination of elemental sulfur in crude oils, the analyst is hampered by the low solubility of the crude oil in the supporting electrolyte, and the procedure of previous investigators was altered where necessary to accommodate this condition. Hall's procedure specifies a supporting electrolyte consisting of 90% methanol, 9.5% pyridine, and 0.5% concentrated hydrochloric acid. This supporting electrolyte is not a satisfactory solvent for crude oils, especially when their elemental sulfur content is low and it is desirable to have about 10% of oil
1745
1746
ANALYTICAL CHEMISTRY
dissolved in the electrolyte. The supporting electrolyte selected t o provide optimum solvency for the crude oils and still give sati5factory polarograms consisted of: 63 27 9 0
0% 0% 570 5%
benzene methanol pyiidine
concentiated hydrochloric acid
It was found that less than 25% methanol in the supporting electrolyte resulted in poor polarographic waves having pronounced maxima. Ethyl alcohol and isopropyl alcohol were tried a5 substitutes for the methanol, but the methanol was the most reliable, yielding consistently good residual-current curves. One shipment of methanol was unusable because a wave produced by an impurity coincided with the elemental sulfur wave Sodium acetate-acetic acid was also tried as a substitute for the pj ridine-pyridinium hydrochloride as a buffer. It was not, however, as satisfactory and decreased the crude-oil solubilitv n ith the mixture of benzene and methanol.
container of alcohol in which it has been stored, wiped carefully, and placed in the cell. The bubbling is continued for 1 minute, then stopped, and the polarogram is obtained using a span of 0.6 volt with no damping. After the polarogram is complete, t h e capillary is removed, washed with a stream of alcohol, and replaced in the alcohol container. The cell is removed from the constant-temperature bath, the sample-containing side of the H-cell is rotated t o discard the sample into a beaker, removed, cleaned with solvent, and dried. The fritted disk of the calomel half-cell is then washed with benzene, followed by alcohol from suitable wash bottles, and dried with filter paper. The joint is lubricated, the sample container replaced, and the cell inserted in the constant-temperature bath for the next determination. Several points in the procedure require careful attention. The elemental sulfur content of samples containing hydrogen sulfid? increases upon exposure to air. Any handling of the crude oil, therefore, or of the dilution after the addition of the crude oil must he done so as t o reduce the time of contact with air to a minimum. Dissolved oxygen in the supporting electrolyte is also difficult to remove completely and if present interferes with the sulfur polarogram. CALlBRATION
Calibration curves of diffusion current LS. weight per cent elemental sulfur were determined for the work Kith both the mercury-pool cell and the H-type calomel cell. Elemental sulfur was dissolved in the electrolyte, and portions of this stock solution were diluted with electrolyte following the procedure used for the crude-oil samples. Figure 2 gives the data ohtained using the H-cell with the calomel electrode. From these data:
I , = 3 i i 0 % sulfur CENTIMETERS
0
2
4
6
and this equation was considered valid between I d = 12.
Zd
=
1.25 and
SCALE
Figure 1. Polarograph Cell
A mercury-pool cell was used in the initial work, and the rrsults were fairly satisfactor>-, but with high-sulfur crude oils there was evidence of blackening oi the mercury surface. .I cell was designed (Figure 1) t,hat redwcd the prohability of the reaction of sulfur with the nwrcur:, 311 LII’ZO cauld be cleaned easily and used conveniently in a constant-Leri~l)erature bsth. This cell is the conventional H-type, with s fritted disk sealed into the inner part of a ground-glass joint in the connecting arm of the H . This construction enables the sample-containing arm of the cell t o be removed for washing without disturbing the calomel halfcell. A second advantage is that the fritted disk can be washed thoroughly by impinging a stream of solvent from a wash bottle upon it and dried by blotting the surface with filter paper. It is necessary t o lubricate the ground joint lightly, as the water in the constant-temperature bath diffuses into the cell rapidly enough to result in an interfering polarogram. Dow-Corning silicone lubricant is satisfactory for this purpose. The calomel half-cell is the conventional saturated calomel electrode, with the esception that 0.1 A‘ lithium chloride is used as the electrolyte and in the preparation of the agar plug. This change, however, was occasioned by another problem where polarograms were being obtained in the range of -2.0 volts and the diffusion of potassium ions into the cell was causing interference. PROCEDURE
The desired amount of crude oil is transferred to a tared bottle M rapidly as possible by means of a pipet. The crude oil is then weighed and the electrolyte, which previously has been bubbled with helium for 10 minutes, is added from a gas-washing bottle. The crude oil plus electrolyte is again weighed, and 5 ml. introduced into t h e polarograph cell and bubbled for 4 minutes with helium. The capillary dropping electrode is removed from a
/’ Figure 2.
Calibration Curve for Use w i t h Calomel Cell
Table 1. Preliminary D e t e r m i n a t i o n o f Elemental Sulfur i n Various Crude Oils
Crude Oil Field Yates Goldsmith Elk Basin Bradford Oregon Basin Hawkins Slaughter Wasson Wilmington Rangely a By bomb. b By polarograph.
State Texas Texas Wyoming Pennsylvania Wyoming Texas Texas Texas California Colorado
Sulfur Wt. % , Totala Elemental6 0.67 2.79 2.17 1.0 0.01 1.95 0.001