Synthetic Detergents from Petroleum - Advances in Chemistry (ACS

---

Synthetic Detergents from Petroleum W. K. GRIESINGER and J. A. NEVISON

Downloaded by COLUMBIA UNIV on July 13, 2012 | http://pubs.acs.org Publication Date: January 1, 1951 | doi: 10.1021/ba-1951-0005.ch028

The Atlantic Refining Co., Philadelphia, Pa.

The development over the past quarter century of both oil- and water-soluble detergent products derived from petroleum is reviewed. Beginning with the by-product mahogany soaps from specialty oil refining, the industry's progress is traced to the tailor-made synthetics of today. Methods of preparation of typical anionic, cationic, and nonionic detergents are described with particular emphasis on the alkyl aryl sulfonates. Product evaluation techniques illustrating the advantages for petroleum synthetics are discussed and a picture of many and varied uses for synthetic detergents is presented.

A typical picture of the display of household cleaners that might be taken today in any supermarket or grocery would show that more than half of the packaged cleaners offered for sale are synthetic detergents. The majority of these contain materials that originated in petroleum. If a similar picture were available for 1932, the year that synthetic detergents began to appear i n volume i n the United States, i t would not reveal a single recognized brand product containing synthetic detergents. The story of this metamorphosis i n consumer preference i n cleaning materials is the history of synthetic detergent chemistry, which is typified b y the development of these materials from petroleum. The incentive to develop a synthetic detergent was a strong one. Although ordinary fatty acid soap has long been, and still is, recognized as a remarkable cleaning agent when used under ideal conditions, its practical use limitations have long been recognized b y the consumer. F o r instance, fatty acid soap is precipitated b y the calcium and magnesium i n hard water. If glasses or clothes are washed i n such water even with normal rinsing, some of the insoluble soap formed usually remains on the glass or textile surfaces as a cloudy gray film. Synthetic detergents have been developed that are completely soluble not only i n hard water but i n sea water as well, and i t is possible to carry out washing operations with these detergents without leaving a residual precipitate on the washed surface. F a t t y acid soaps are invariably alkaline, having a p H of 9 or higher i n dilute solutions. This alkalinity is particularly undesirable i n the washing of fine silks and woolens, where it can accentuate fabric shrinkage or color fading. Synthetic detergents have been developed that are neutral and have no more effect than plain water on fine fabrics. A t p H values below 6, fatty acid soaps are completely hydrolyzed to insoluble fatty acids which have no detergent action. B u t some cleaning operations, specifically i n the dairy industry, are best carried out with acid baths. Synthetic detergents have been developed which are stable under these conditions (22). I n the synthesis of chemical compounds designed to overcome the disadvantages of soap, but maintaining a l l of soap's good qualities, one of the biggest problems has been the development of reliable laboratory evaluation methods for measuring detergency and 324

In PROGRESS IN PETROLEUM TECHNOLOGY; Advances in Chemistry; American Chemical Society: Washington, DC, 1951.

325

GRIESINGER AND NEVISON—SYNTHETIC DETERGENTS FROM PETROLEUM

lathering. These laboratory methods lean more to the practical than to the scientific side of the problem. I n measuring the detergency of a material with respect to textiles, a typical laboratory evaluation method includes: soiling a standardized textile swatch with a soil comprising fatty and mineral oils and carbon black; agitating this swatch i n the detergent solution under test i n a miniature washing machine called a launderometer; then measuring by means of light reflectance the amount of carbon black removed during the operation (10, 11, 14). A typical detergency pattern can then be obtained by plotting the per cent of detergency against the detergent concentration and comparing the slope of the curve with that of a detergent of proved performance such as tallow soap (Figure 1).

1

Downloaded by COLUMBIA UNIV on July 13, 2012 | http://pubs.acs.org Publication Date: January 1, 1951 | doi: 10.1021/ba-1951-0005.ch028

9 0 e

WEIGHT % CONCENTRATION

Figure 1.

Colton Detergency

For measuring lather and lather stability, a simple means for recording the foam height resulting from agitation of the detergent solution under controlled conditions—shaking in a graduated cylinder or dropping through a standardized column—has been used (2, 20). More faith, however, is put i n practical tests, such as one i n which dishes are soiled with a mixture of egg and bacon grease and then washed i n a given volume of detergent solution. The number of dishes that can be washed before all the lather disappears from the dishpan is a measure of lather stability. T o the skilled physical chemist these evaluation procedures may appear crude, but a rather good correlation has been obtained between laboratory evaluations and performance as measured by the consumer under use conditions. N o serious marketer of deterIn PROGRESS IN PETROLEUM TECHNOLOGY; Advances in Chemistry; American Chemical Society: Washington, DC, 1951.

ADVANCES IN CHEMISTRY SERIES

Downloaded by COLUMBIA UNIV on July 13, 2012 | http://pubs.acs.org Publication Date: January 1, 1951 | doi: 10.1021/ba-1951-0005.ch028

326

gents, however, offers new products for sale until he has first laboratory tested them and then obtained favorable consumer reaction i n a blind market survey. I n this the test material and a product of known value are compared by a cross section of the consumer public. This expensive and rather cumbersome method of proving the value of cleaning compounds has probably retarded the progress of synthetic detergents more than any other single factor, including the quality of the products offered and the desire of the public to accept them. A s far back as 1932, the general structure of a good detergent was recognized and the intervening years have served primarily to contribute to the quality improvement of basic materials. Color, odor, wetting, and washing efficiency have been continually bettered but the theoretical concepts of a good detergent have not changed. The most effective detergents are characterized b y a molecular structure which is bifunctional—one portion consisting of a hydrophobic hydrocarbon radical and the other portion consisting of either an inorganic ionizable group or a polar group which is hydrophylic (28). Such chemical structures, illustrated i n the formulas of Table I , are common with the broader class of products termed surface-active agents; for, b y virtue of their structures, solute molecules concentrate at interfaces altering surface effects as evidenced by changes i n surface tension, wetting speed, detergency, and foaming characteristics. Detergents constitute a relatively limited molecular weight group i n this class of products, as requisite properties are obtained only when a precise balance between the two portions of the molecule exists. Such a balance usually results i n borderline solubility i n both aqueous and organic media. Numerous materials of varying chemical structure meeting these requirements can be made from petroleum.

Alkyl Aryl Sulfonates The a l k y l a r y l sulfonates, specifically those having a molecular weight and structure approaching dodecyl benzene sulfonate, were among the first and have grown to be the major detergent products produced from petroleum today (6). The development of the better detergent products of this approximate composition has followed the expedient course of synthesizing and testing those homologous and isomeric mixtures of compounds which may be prepared from available low-cost raw materials b y practical reaction techniques. The preparation and sifting of individual isomeric products as a means to finding optimum structures have received but limited attention—and, indeed, would be an impractical approach, as there are 38,636 possible isomeric CigHso a l k y l benzenes, 355 of which are monoalkyl benzene isomers (7). A s the more expedient syntheses have evolved, however, and improved analytical methods have developed, increased attention has been given to correlations between molecular weight and structure of process reactants and final product performance characteristics.

Green Acid Sulfonates The first a l k y l a r y l sulfonates produced i n this country were those which resulted from the sulfuric acid refinement of kerosene, gas oil, and lubricating oils and were initially

Table I.

Ionic Classification of Detergents

Tallow soap or sodium stéarate

Anion CnHseCOO-

Cation Na

Sodium dodecyl benzene sulfonate Alkyl naphthal trimethyl ammonium chloride

Poly ethenoxy ether of iso-octyl pheno

In PROGRESS IN PETROLEUM TECHNOLOGY; Advances in Chemistry; American Chemical Society: Washington, DC, 1951.

+

327

GRIESINGER AND NEVISON—SYNTHETIC DETERGENTS FROM PETROLEUM

noted for their nuisance value alone (5). A s their occurrence and refinement became better understood, i n several cases they have evolved to useful by-products and finally to primary products. The water-soluble sulfonic acids resulting from the acid treatment of oils boiling under 550° F . — k n o w n as "green acids" i n the trade—upon extraction from the spent sulfonation acid sludge with a suitable solvent, gave water-soluble salts of limited commercial utility. A choice oil for producing high yields of a green acid type product would be a refractory cracked gas oil from a relatively aromatic crude such as Lea-Ward, and would have the inspections shown i n Table I I . Table II.

Inspections on Cracked Gas Oils from Lea-Ward Crude %

Downloaded by COLUMBIA UNIV on July 13, 2012 | http://pubs.acs.org Publication Date: January 1, 1951 | doi: 10.1021/ba-1951-0005.ch028

A S T M distillation

A P I gravity Components Olefins, % Naphthenes, etc., %

°F.

0 248 10 403 50 466 00 526 95 564 E.p. 590 26.2 20 to 22% polyalkyl aromatics, some polynuclear 20 to 25 52 to 60

Aromatics i n this oil can be converted almost quantitatively to sulfonates b y controlled sulfonation with 9 8 % sulfuric acid at about 150° F . The sulfonic acids are soluble in the spent acid and must be extracted with a nonreactive solvent. Similar treating procedures can be used on more concentrated aromatic materials such as the polyalkylated benzenes which are formed as heavy by-products i n the manufacture of cumene from propylene and benzene (18). Principal use of these products as detergents occurred during W o r l d W a r I I when they were used with fatty acid soaps i n formulating army all-purpose detergent bars (25). Their low molecular weight and high water solubility limited their value for use alone i n efficient detergent formulations.

Oil-Soluble Sulfonates The oil-soluble sulfonates resulting from the acid treatment of oil boiling above 550° F . are typified b y the mahogany or white oil sulfonates which for many years were the unwanted by-products of white oil manufacture, but which i n view of recent industrial applications have reversed positions with the former primary product. Mahogany sulfonic acids are most readily formed b y the action of fuming sulfuric acid on a naphthenic-type petroleum distillate, preferably one having a Saybolt Universal viscosity at 100° F . of about 800 to 900 seconds and an A P I gravity of about 22. Following acid treatment of such a distillate with approximately 40 volume % of acid, the acid or sour o i l is simultaneously neutralized and extracted with alcoholic caustic. T h e mahogany sulfonates which are soluble i n the alcohol extract are then freed of excess sodium sulfate and caustic b y drying and redissolving i n a low boiling hydrocarbon (5). These products find their main applications as emulsifiers, rust preventives, and detergents for use i n motor oil. F o r the latter application, salts other than sodium are preferred and the conversion can be made either b y double decomposition or b y neutralization of the original sulfonic acids with the desired base (9). The detergent effect of barium and calcium sulfonates i n heavy-duty lubricating oils, particularly those used i n Diesel equipment, is evidenced b y the striking decrease i n engine deposits and ring fouling resulting from their use (8). While synthetic oil-soluble sulfonates have also been made b y reacting high molecular weight olefins or chlorinated wax with selected aromatics and sulfonating the resulting compound, the oil detergent properties of these materials have been reported to be inferior to those made b y conventional acid-treating procedures on petroleum distillates. In PROGRESS IN PETROLEUM TECHNOLOGY; Advances in Chemistry; American Chemical Society: Washington, DC, 1951.

328

ADVANCES IN CHEMISTRY SERIES

Keryl Benzene Sulfonates I n the early thirties, at about the same time that sulfonated petroleum hydrocarbons were being rigorously explored for utility i n the surface-active field thefirstkeryl benzene sulfonates were produced. These products, prepared from highly paraffinic fractions averaging about fourteen carbon atoms per molecule, are synthesized according to the reactions: RH

+C1

*RC1

2

Kerosene hydrocarbons

+HCi

(1A)

K e r y l monochlorides AICI3

Downloaded by COLUMBIA UNIV on July 13, 2012 | http://pubs.acs.org Publication Date: January 1, 1951 | doi: 10.1021/ba-1951-0005.ch028

RC1 + benzene

R - ^ ~ ^ >

+ H S0 2

R-~ R — < f > + HC1 140-180° F . \ = / K e r y l benzenes V-SOaH + H 0

> R— R— + H S 0 2

4

R— R - ^

V-SO3H + H 0

> R— RS0 C1 + HC1

(3A)

> RS0 Na + NaCl

(3B)

2

Paraffin preferred RS0 C1 + 2NaOH

Downloaded by COLUMBIA UNIV on July 13, 2012 | http://pubs.acs.org Publication Date: January 1, 1951 | doi: 10.1021/ba-1951-0005.ch028

2

8

F o r the paraffin, a highly acid-treated o i l having a Saybolt Universal viscosity of 40 to 50 seconds at 100° F . is preferred. Reaction is rapid at 200° F . using about 1.7 parts b y weight of sulfur dioxide and 0.9 part b y weight of chlorine for each part of paraffin. The resulting sulfonyl chloride is neutralized with sodium hydroxide and separated from unreacted oil prior to drying. The detergent properties of the a l k y l sulfonates approach those of the a l k y l aryl sul­ fonates, but the commercial processing problems have placed these materials at a n eco­ nomic disadvantage. Somewhat analogous to the Reed process is a recent development i n which an olefin is reacted with nitrosyl chloride to form a nitrosyl halide, which i n turn is converted to a mixture of complex sulfonates b y reaction with sodium sulfite (4). These materials are reported to be excellent metal cleaning detergents.

Alkyl Sulfates Another class of detergents that merits attention is represented b y the a l k y l sulfates. These materials have been prepared b y the direct sulfation of high boiling olefins resulting from the cracking of paraffin wax or the sulfation of alcohols derived from petroleum. The direct sulfation of wax olefins has been perfected i n Europe but has not been commercialized to any extent i n the United States. M o r e work has been done i n this country on alcohols that have been prepared b y Fischer-Tropsch syntheses, oxo process reactions, and reduction of the fatty acid mixtures obtained b y paraffin wax oxidation. I n most instances these alcohols as well as the olefins have been branched chain or second­ ary products, both of which have been reported to give inferior detergent a n d sudsing properties (18, 24).

Nonionics There is a whole class of synthetic detergents, originating partly or wholly from petroleum, characterized b y the fact that the detergents do not ionize i n water. These compounds are high molecular weight polyglycol ethers and esters which usually are pre­ pared b y reacting ethylene or propylene oxide with a polar organic compound having a replaceable hydrogen as i n an O H , S H , or N H group (12). B y way of example, iso-octyl phenol from diisobutylene and phenol reacts quantita­ tively with ethylene oxide i n the presence of an alkaline catalyst to add any desired n u m ­ ber of (CJEUO) groups as follows:

Ο C H OH + 8

1 7

xClCcH

2

—> C H 8

1 7

0(C H 0) -iC H40H 2

4

a

2

(4)

F o r good detergency the length of the ethenoxy chain should carefully balance the hydrophobic part of the molecule. I n the above example, from eight to ten such groups appear to be optimum. T h e ethylene oxide for these products comes from petroleum ethylene, either b y direct oxidation or decomposition of ethylene chlorohydrin (17). I n addition to the alkylated phenol mentioned, the hydrophobic part of the molecule may be derived from the fatty acids resulting from wax oxidation, and mercaptans from

In PROGRESS IN PETROLEUM TECHNOLOGY; Advances in Chemistry; American Chemical Society: Washington, DC, 1951.

331

GRIESINGER AND NEVISON—SYNTHETIC DETERGENTS FROM PETROLEUM

the reaction of high molecular weight olefins with hydrogen sulfide or Oxo process alcohols (12,19).

Nonionic detergents are usually viscous liquids with excellent solubility i n all kinds of water. They are particularly good detergents for wool and have found their widest acceptance i n this branch of the textile industry.

Cationics The final class of water-soluble synthetic detergents derived from petroleum are the cationics. These products, when ionized, have the hydrophobic portion of the molecule i n the cation. They are of interest principally because of the bactericidal properties they exhibit. Their performance as detergents is poor. The principal products marketed to date have been derived from normal long-chain amines, but procedures for their preparation from petroleum hydrocarbons (1) have been described as follows: CH 0 Downloaded by COLUMBIA UNIV on July 13, 2012 | http://pubs.acs.org Publication Date: January 1, 1951 | doi: 10.1021/ba-1951-0005.ch028

2

ZnCi

CH C1

2

A l k y l naphthalene CH C1 + N ( C H ) 2

3

(5A)

2

*R-

^1 + H C l

3

CH3

CH - - N — C H

Refluxing MeOH

CH3

2

I

R-

3

(5B)

CI

A l k y l napthal trimethyl ammonium chloride

Detergent Forms Synthetic detergents are marketed as liquids, slurries, drum-dried flakes and spraydried beads, the latter being the form most frequently appearing i n consumer packages. L i q u i d detergents are used primarily for dishwashing or shampoos and usually are composed of highly water-soluble materials such as the nonionics, the lower molecular weight a l k y l aryl sulfonates, the organic amine sulfonates, or mixtures of the same. Active product concentrations i n water or an alcohol solvent will vary from 5 to 9 5 % . D r u m dried flakes find their biggest outlet i n industrial detergent mixtures where the original 30 to 9 0 % active flakes are frequently dry-mixed with alkali builders for specific cleaning applications. The spray-dried beads are formed i n a spray tower b y blending a host of additives or builders with highly active detergent slurries and drying to homogeneous beads. F r o m the consumer viewpoint there are two important and distinct products which are referred to as light- and heavy-duty detergents. T h e light-duty products, recommended for use on dishes and fine fabrics, usually have the composition shown i n Table I V . Table IV. Active detergent Tripoiyphospbate Fluorescent dye Sodium sulfate

Light Duty Synthetic Detergent 20-35 5-10 0-0.5 75-55

Heavy-duty products, designed specifically for the cleaning of cottons, are illustrated i n Table V .

Production Trends The production trends (3) shown i n Figure 2 are for all synthetic detergents. N o split-up is available for petroleum-based materials alone, but the indications are that better than 5 0 % of the products fall into this category. A s can be seen from the curve, the In PROGRESS IN PETROLEUM TECHNOLOGY; Advances in Chemistry; American Chemical Society: Washington, DC, 1951.

332

ADVANCES IN CHEMISTRY SERIES Table V.

Heavy Duty Synthetic Detergent %

Downloaded by COLUMBIA UNIV on July 13, 2012 | http://pubs.acs.org Publication Date: January 1, 1951 | doi: 10.1021/ba-1951-0005.ch028

Active detergent Tripolyphosphate Fluorescent dye Carboxy methylcellulose Sodium sulfate

25-40 25-50 0.5-1.0 1.0-4.0 49-5.0

industry is still growing and can be expected to continue, u p to perhaps 3 billion pounds per year. Petroleum-based products should get a lion's share of this market, provided that the raw materials—particularly benzene and olefins of the right molecular weight and struc­ ture^—can be made available at prices that w i l l permit competition with synthetics a n d improved soaps based on products from agricultural sources.

Literature Cited (1) Adams, R., "Organic Reactions," Vol. I, New York, John Wiley & Sons (1942). (2) American Society for Testing Materials, Preprint 91, Report of Committee D-12 on Soaps and Other Detergents, p. 5 (1950). (3) Anon., Soap and Glycerine Producers Association, "Statistics." (4) Beckham, L. J., U. S. Patent 2,265,993 (1941). (5) Ellis, C., "Chemistry of Petroleum Derivatives," Vol. II, pp. 1069-81, New York, Reinhold Publishing Corp., 1937. (6) Flett, L . H . , U . S. Patent 2,283,199 (1942). (7) Francis, A . W., Chem. Rev., 42, 107 (1948). (8) Geniesse, J. C., and Schreiber, W., Oil Gas J., 49, 302 (1951). (9) Griesinger, W. K., U . S. Patents 2,361,476 (1944), and 2,406,763 (1946). (10) Griesinger, W. K., and Nevison, J . Α., J. Am. Oil Chemists' Soc., 27, 88-90 (1950). (11) Harris, J. C., American Society for Testing Materials, Bull. 140, 141 (May and August 1946). (12) Hoyt, L . F., Office of Technical Services, Dept. of Commerce, Washington, D . C., PB-3868 (1945). (13) James, J. H., U. S. Patent2,341,218(1944).

In PROGRESS IN PETROLEUM TECHNOLOGY; Advances in Chemistry; American Chemical Society: Washington, DC, 1951.

GRIESINGER AND NEVISON—SYNTHETIC DETERGENTS FROM PETROLEUM

333

Lambert, J . M . , and Sanders, H . L . , Ind. Eng. Chem., 42, 1388 (1950). Lewis, A . H . , and Ettling, A. C., U . S. Patents 2,477,382-3 (1949). Lisk, G. F., Ind. Eng. Chem., 40, 1671 (1948); 41, 1923 (1949); 42, 1746 (1950). McClellan, P. P., Ibid., 42, 2402 (1950). Nevison, J . Α., Can. Patent 453,974 (1949). Nevison, J . Α., and Griesinger, W. K., U . S. Patent 2,542,697 (1951). Ross, J., and Miles, G . D., Oil & Soap, 18, 99 (May 1941). Rossini, F . D., and Mair, B. J., Proc. 22nd Annual Meeting, API, Sect. III, 22, 7-12 (1941). Scales, F . M . , Food Industries, 14, No. 4, 53 (1942). Schwartz, A . M . , and Perry, J . W., "Surface Active Agents," New York, Interscience Publishers,Inc., 1949. (24) Tulleners, U . S. Patent 2,078,516 (1937). (25) U. S. Army Quartermaster Corps, Tentative Specification, O Q M G No. 100A (Jan. 18, 1944). (26) Zimmerschied, W. J., etal.,Ind. Eng. Chem., 42, 1300 (1950). (14) (15) (16) (17) (18) (19) (20) (21) (22) (23)

Downloaded by COLUMBIA UNIV on July 13, 2012 | http://pubs.acs.org Publication Date: January 1, 1951 | doi: 10.1021/ba-1951-0005.ch028

RECEIVED April 22, 1951.

In PROGRESS IN PETROLEUM TECHNOLOGY; Advances in Chemistry; American Chemical Society: Washington, DC, 1951.