Viscosity Index Improvers for Lubricating Oils. Polyvinyl Esters ofLong

fore other methods have been sought. Use of thickening agents in lubricating oils has been especially successful. In particular, addition of certain o...
1 downloads 0 Views 390KB Size
Viscosity Index Improvers for Lubricating Oils POLYVINYL ESTERS OF L O N G - C H A I N FATTY ACIDS WILLIAM S. PORT, JAMES W. O’BRIEN, JOHN E. HANSEN, AND DANIEL SWERN Kastern Regional Research Laboratory, Philadelphiu 18, Pa.

molecules, with a resulting industrial Polymers and copolymers of vinyl esters of the longincrease in intrinsic viscosity. and military use of machain fatty acids are of potential commercial interest beIn a “good” solvent also, a chines operating through wide cause the monomers can he easily prepared in good yield t e m p e r a t u r e rise tends t o tempprature ranges has defrom inexpensive and readily available raw materials. diminish the intra- and interveloped great interest in As part of a general program of preparation and evaluamolecular bonding, but since lubricating oil& having imtion, these polymers and copolymers were tested as visthe polymer chains are alp r o v p d viscosity-temperacosity index iniprovers for lubricating oils. ready uncoiled, the reduction ture characteristics. The imPolyvinyl palmitate, polyvinyl caprylate, and copolymers in bond strength is necessaril~. provement in the viscosity of vinyl palmitate with vinyl acetate are effective viscosit? Plight. A temperature rise, index which can be obtained index improvers for lubricating oils. The improvement liowever, causes a greatel, from increased refbiny and caused hy the copolymers increases with increased vinyl from mixing of pet,rolcuni w e a k e n i n g of the strong acetate content. The soliihility of the copolymers in solvent-to-polymer b o n d b , stocka is limited, and thereliihricating oils decreases as the vinyl acetate content allowing increased intra- and fore other methods have been increases; an optimum effect is obtained with a copolymer intcrmolecular i n t e r a c t i o n , Bought. Use of thickening containing about 25 mole c/c of vinyl palmitate. and, therefore, the net effcct agents in lubricating oils has In general, the viscositj index-improving effect of a is a coiling up of thc polymer been e,qxcially successful. In polymer or copolymer in a lubricating oil is enhanced as molecules with a resulting departicular, addition of certain the solubility of the poly-mer decreases and its concencrease in intrinsic viscosity. oil-soluble polymers to a tration increases. Experimentally, these authore basc oil has greatly improved used a mixture of a solvent its viscosity index. Among suitable polymers in use todwy are polyhutenes (R. 1 4 J. ] ) n l ~ . and a precipitant a~ the “pooi.“ .solvent and the pure solvent as mrlhncrylates (8, 26)?and ~.’olyalkystyi,eiir.s (;). the “good” solvent. Evans and Young ( 4 ) purforimd csperinicnts similar to thosr VISCOSITY OF HIGH POLYMER SOLUTIONS of Alfrey e! nl. (1 ), but they studied polybutenes in some solvents ailid in lubricating oils containing varying amounts of oasr 01’ oil. T l i c , viscosity of a high-polymc!r solution ( 1 , 1 8 ) depends on phthalic acid, and glycol derivatives a,s the nonsolvtmts f(Jr thr, t h size of the polymer, the extent-of the intra- and interniolecular polynwrs, :ind obtained similar results. interactions of the polymer rnoleculos, the extent of the interacS C;K EASED

tion l)ct,weenthe polgmc-r molecules and the solvc~it,the concentration of the polymer and the natuw of the solvent. In general, because higher molecular volume acts to impede liquid flow, R polymer of a high degree of pol3.-rui.rizrttion gives a solution of higher viscosity than docs a low molecular \vcsight. po1ymc.r. If the int,ramolecular forces are high and the degree of solvation is low-that is, a “poor” ~olvcnt-thr internal :ittr:tcting foi~ces cauw the polymer molecules to curl up in compact forni; this decrease6 the effective volume :tnd surfacc. xrc:~of contitct anti gives solutions of low viscosity. In polymer solutions in which tlw iiitra- :inti iiitermo1ec:ulai. forces are low and solvation forces arc liigh---a “good” solviwt -each polymer molecule: is sheathed 1)s solvent molccules :tiid tends to uncurl into long chains; this increases thr effective volume and t,he surface area of contact and givcs soliltions of higher viscosity than in a “poor” solvent. . The effect of temperature on the viscosities of solutions of Iiigli pulyniers has bwn discussed by Alfrey, Rartovics, and Miirk ( 1 1, who studied polystyrene in various solvent mixtures. T h r ~ , showed that in a ‘(poor” solvent the intrinsic viscosity of flexible chain polymers increases II it,h tempcratuvc whereas in a “good” solvent it decreases, and the-v proposed x mechanism to explain this phenomenon. I J a~ “poor” solvent, an increase in teniperature diminishes the ptreiigth of thr intra- and intermolecw1:tr. bonding: this increases the surfarc itnd vo1111iieof thv polynicr

AI’PLICATION OF ‘rAEORY T O YlSCOSITY INDEX IMPHO) EHS

Thiw’ t1it.oric.s of‘ polymer solutio~ish;ive a direct bearing on t,he use of polymws as viscosity indcsx improvers. Three conclusions may be tirnAn. First, as ib experimentally well known, polymers of a high dogrec of polyrnc~rizationraise the viscosity index more than dow itn equal conc*mt,rationof lower niolecular weight. polymers. Thi- method of augmenting the effect of a given polymer a s a viscosit y inticsu improver suffers because high mnltc-ul:tr weight p l y i i i ~ r bmay he unstable to shearing force,? during use, thus 1i:niting tlie degree of polymerization of polymers that can be utilized. Second, a polymer for which an oil is a “pour” solvent \vi11 be a. better viscosity index improvcr than one for n11ic.hthe oil is a “good” solvont. Furthrr, the greater ihe cwncent,i,:ition of polymer in a, “poor” solvent, the greater will t)e the viscosity index improvement. Third, as was also concluded by Evans and Young ($), any method by which the sohI)ility of a polymer in a lubricating oil may be decreased will incwtm~t lie viscosity index-improving effrct of the polymer. The solvent powers of t,lie lubricating oils must necessarily br iwiisidcred as essentiaily constant, because addition of large umouiits of nonsolvents, such as those used by Evans and Young ( h ) , ivould greatly increase the cost of the oil. However, if two monomers are available, and the homopolymer of onc is highly aoluhle and that of thf, other is insoluble in pet,roleum oils, t l i c ~ r i

210s

INDUSTRIAL AND ENGINEERING CHEMISTRY

2106

Vol. 43, No. 4

the potential advantage of a lower molecular weight for the same degree of polymerization and lower costs than do the homopolymers. PREPARATION OF POLYMERS AND COPOLYMERS

1

I

,L

OIL

./

0

I

A

I

T

OIL 8 7

2 3 4 POLYVINYL PALMITATE, %

Figure 1. Viscosity Index of Lubricating Oils C o n t a i n i n g Polyvinyl P a l m i t a t e This paper describes the use of relatively new polymers, the polyvinyl esters of the long chain fatty acids, as viscosity index improvers. [Wulff and Breuers (16) and I. G. Farbenindustrie (6) mentioned the use of some polyvinyl esters in lubricating oils but gave few data.] Because the vinyl esters are easily made in good yield (11, IS) from inexpemive and readily available raw materials (acetylene or vinyl acetate and fatty acids), they are potentially low cost monomers. Furthermore, these monomers can be polymerized (10)and copolymerized readily to give products of wide applicability. Therefore, studies of the uses of these pollmers and copolymers may be of industrial importance. The polymers used in this study were polyvinyl caprylate and polyvinyl palmitate. These were selected as representative of a liquid and a solid polymer and of a moderately long and a longchain fatty acid. The copolymer syPtem chosen for study was vinyl acetate and vinyl palmitate. Polyvinyl acetate is insoluble in petroleum solvents and represents the insolubilizing portion of the copolymer. Vinyl palmitate was chosen as the comonomer because the resulting average chain length of the acyl radicals in the copolymer was approximately that in polyvinyl caprylate, which gave good results. In addition t o the possibility of imparting significant increases to the viscosity index, the copolymers also have

The polymeric materials used in this study were prepared in solution to control the degree of polymerization. For a degree of polymerization of about 400, a solution of 1 mole of monomer and 0.5 mole 7 0 of benzoyl peroxide as the initiator in 2.5 moles of benzene was heated for about 7 hours a t 75" C. in a stream of nitrogen. Conversion to polymer was 60 to 70%. The palmitate was freed of monomer by two precipitations with acetone from benzene solutions; methanol was employed as the precipitant for the caprylate. Because the palmitate precipitated as a solid and did not occlude solvent, it was readily freed of adhering solvent by air and vacuum drying. The caprylate was freed of solvent by allowing its benzene solution to evaporate to dryness first at room temperature and then in a vacuum oven a t 100" C. As determined by the light-scattering technique, the caprylate had a weight average molecular weight of 73,000, and the palmitate, 119,OOO. To obtain a polymer having a degree of polymerization of about 100, ethylbenzene-a more efficient chain transfer agentwas used. Polyvinyl caprylate having a molecular weight of 20,000 was prepared by heating the monomer with an equimolar amount of ethylbenzene and 0.5 mole % of benzoyl peroxide for 12 hours a t 75" C. The purification and isolation techniques were the same as for the higher molecular weight polymer. T h e yield of polymer was 60%. Copolymers of vinyl palmitate and vinyl acetate were prepared in 3: 1, 1:1, and 1 :3 molar proportions by heating the a p propriate monomer mixtures and 0.5 mole % of benzoyl peroxide in 2.5 moles of benzene for 7 hours a t 70" to 73" C. The copolymers were freed of monomers by three precipitations with methanol from benzene solution. Solvent was removed by the frozen-benzene technique (6). The yields of the copolymers were about 60%. The weight average molecular weights of the copolymers containing 75 and 50 mole % of vinyl palmitate were 95,000 and 92,000, respectively. PREPARATION OF T E S T SOLUTIONS

Table I shows the viscosity and pour point properties of the untreated oils used. The viscosity indexes were calculated from t,he tables of Dean and Davis (9). TABLEI. EFFECTOF POLYVINYL ESTERSON VISCOSITY AND

POUR POINT

Polyvinyl Ester (5% Solution) None CiGFy~ate prylate

Palmitate-acetate

Approx. Degree of Polymerization

*..

400 100 400 400 600

...

..

. None Caprylate 400 Caprylate 100 Palmitate 400 Palmitate-acetate (75/25) 400 Palmitate-acetate (50/50) 500

Viscosity Index (3)

Pour Point

5.17 9.01 5.81 8.65 7.97 8.61 6.47

90.2 131.9 112.4 123.7 130.6 132.5 136.5

25 25.

10.38 14.98 12.26 16.12 14.32 14.05

79.1 113.1 86.0 98.5 106.6 104.7

01 5.

7.24 13.31 9.47 12.34 12.12 12.07

54.5 98.9 78.9 96.7 96.0 95.8

Viscosity*cS. 100" F. 210' F. Oil A a 33.13 57.61 45.24 59.15 49.25 54.08 36.07

(2)

..

40 40'

Oil B b 110.8 139.7 135.9 182.9 141.1 140.3 Oil CC 71.50 137.2 96.18 125.7 122.6 122.1

None Caprylate 400 Caprylate 100 Palmitate 400 Palmitate-acetate (75/25) 400 Palmitate-acetate ( 5 0 / 5 0 ) 500 a Pennsylvania 150 neutrul. b sol vent-refined nlid-continent neutral. C .Acid-refined mid-continent neutral.

45. 45.

25 25

40 35

..

September 1951

INDUSTRIAL A N D E N G I N E E R I N G CHEMISTRY

The polyvinyl caprylate and the polyvinyl palmitate having titagrees of polymerization of about 400 were used t o inrilie solutio1i.c containing 0.5, 1, 2, 3, and 5yo of polymer in each of thc. threc t,est oils. Because the rate of viscosity index improvcmrnt ivitli 1,iLspec.tto concentration appeared to diminish a t about jC; (Figures 1 and 2), only solutions containing 5% of polymor \ ~ e t t niadc from the lower molecular weight, polyvinyl caprylate and from the copolymers. The solubility of the 25375 copol>.mer (pa1mitatc:acetate) \vas less than 3yoin oils B and C, and they(.[ore the copolymer )vas not tested in these oils. The 5 7 , polyvinyl palmitate and the ;ice 25 : 75 copolymer solution t)ccanit~ Iii~tc~rogeneous o n long standing at room temperature.

:

O

4

2107

~-

polyviny) Viscosity, Ce. Initial _ _ _ _l_~ _ i n_a l " bkter, 3% Concn. 100' I:. 210' F. 100' F. 210' F. Control (base oil 13) 110 8 10.38 118.3 10.73 (:aprylatr 130 7 14.98 117.7 13.29 l'alniitateacetate (75/23) 1a1.1 14.3% 138.7 13.44 a . i f t e r heating a t 300° 1 .: for 100 hours.

_Viscosity ___ Index _~~~ ~

Initial 79.1

113,l 106.6

Final"

77 2

90.9

81.5

The ch:inges in viscosity index a n d in viscosity st 100" F. in oil A produced by the copolymers as compared v;ith the changes produccd b y polyvinyl palmitate are in accord with the theory of viscosity index improvement by polymers. The viscosity index increases and the viscosity a t 100' F. decreases as the acetyl content of the copolymer is incrrased. In oil B, too, the viscosity indcs improvement caused by the copolymers is greater than that cawed by polyvinyl palmitate, but the viscosity indcs improvement is not correlated with acet,yl content. In oil C, the copolymers did not c:iuse any greater increase in the viscosity indcx than did polyvinyl palmitate, but did cause a decrease in the viscosity a t 100" F. Table I1 shows the effect of heating one of the test oils and solutions of one of the copolymers and one of the homopolymers in the same oil for 100 hours a t 300" F. in air. This test is intended t o simulate the useful life of a lubricating oil. At the end of the heat,ing period, the viscosity index decreased in all cases but the viscosit,y index of the oils containing additive still was significantly higher than that of t.he uiit.reated oil.

I

ACKNOWLEDGMENT

70y 1 60

0

i 2 3 4. POLYVINYL CAPRYLATE, %

5

Figure 2. Viscosity Index of Lubricating Oils Containing Polyvinyl Caprylate

RESULTS A N D DISCUSSION

l'igures 1 and 2 show the effect of increasing coiiccntratioiis of t h e two polymctrs having a degree of polymerization of about -100 O I I the viscosity indcs of the three base oils. It may be seen that illcreasing the concentration of the polymers increases the viscosity indpx and that the rate of improvement declines a t about 5 yo 'I'nhle I gives the viscosity, viscosity index, and pour point

:lata for the solutions containing 570 of all the polymers tested. ThPRe data demonstrate that both polyvinyl caprylate and poly-

vinyl palmitate having a degree of polymerization of about 400 :we effective lis viscosity index improvers. When used in 5Y0 conrentamtion,cacti polymer improved the viscosity properties of thc tnse oils suffiriently to incrense the S.A.E. rating from 10 to 20-20 W, from 30 to 40, and frhm no rating to S.A.E. 30. Of the two polymers, polyvinyl caprylate showed the better properties, b r w u s c it had a grwter effect on the viscosity indes and did not raise the pour point. Polyvinyl caprylate having a degree of polymerization of :Ltmut 100 did not exhibit any appreciable visrohity indrx improving qualities whrn used in 5% concentration.

The authors thank Phoebe L. Proctor for assistance in determining the molecular weights of the polymers, and J. C. Geniesse of Atlantic Refining Co. for samples af lubricating oils. LITERATURE CITED

Alfrey, T.. Bartovics, -4..and Mark, 11..J. Am. Chem. Soc., 64 1557 (1942).

Am. So?. for Testing Materials, Philadelphia ASThI Standards 1949, Pczrt 5 , pp. 73&9 (1949). Dean. E. JT-., and Davis. G. H. B., Chem. & Met,. Eng., 36, So. 10, 618 (1929). Evans. H. C., and k'oung. D. K., IIUD. E s o . CHEM.,39, 1676 (1947). I. G . I'arbenindustiie, -4.-G., Ger. Patent 600,722 (Feb. 15, 1936). Lewis, F. iM.,and Mayo, F. R., IND.ENG.CHEX, h 4 1 , . E D . , 17, 134 (1945). IIikeska, L. A,, and Fulton, S. C., U. S. Patent 2,072,120 (Xfarch 2, 1937). S e h e r , €1. T., and Hollander, C. S., Ibid., 2,114,233 (1938). Otto, AI., and Mueller-Cunardi, A I . , I b i d . . 2,130,507 (Sept,. 20, 1935). P o r t , W. S., Hansen, J. E., Jordan, E. I?., Diet., T. J.. and Swern, D., J . Polymer Sei., 6 , i n press. Reppe, W., Ger. Patent 588,352 ( S o v . 15, 1933). Schmidt, A . S., and Mm-lies. C . A , , "I'rinciples of High Polymer Theory and Practice." pp. 77-82, 247-50, Piew York, McGraw-Hill 130ck Co., 19-1-3. Swern, D., and .Jordan, E. F.,J . .4m. C h e m . SOC.,70, 2334 (1946). Thomas, It. VI., Zimmer, J. C., Turner, L. B., Roson, R., and Frohlich, I-'.K.. 1x0. E N G .CnEw., 32,299 (1940). Van I l o r n e , TI'. L., Ibid., 41, 952 (1949). W'uiff, C., and Iireuers, M., L.S.Patcnt 2,020,714 ( S o v . 12, 1935).

RECEIVBD January 11, 1951. ?resented before t h e Petroleum Section at the ~ ~ e e t i n g . - i n - l \ l i n i a t uof r e the Philadelphia Section, AXERICAN CHEMICAL SOCIETY, January 18, 1951. Report of a study in which certain phases were carried on under the Research a n d Marketing Act of 1916. This paper is V in the series "I'olymeriaable Derivatives of Long-chain F a t t y Acids." Paper I V is (IO).