Quantitative thin layer chromatography of lipids

JOURNAL

388

QUANTITATIVE

M. LOUISE

THIN

CUZNER

AND

DejJarlme& of Bioc7temistl#y, WC: 2 (Great Britain) (Rcccivccl Dccembcr

LAYER

A.

CIIROMATOGRAPEIY

OF CEIROMATOGRAPNY

OF LIPIDS

N. DAVISON

Citaving

Cross

Uoq5ilaZ

Medical

School,

62-65

Cltandos

Place,

London,

soth, IgGG)

Thin layer chromatography is routinely used for the separation of all classes of lipids and has now been variously aclopted to provide the basis for quantitative analysis of separated lipidi. Thus, CLAUSEN, Lou AND ANDERSEN~ and SIAKOTOS AND ROUSERS have described methods of analysis based on densitometric measurement of spots visualised under standard conditions. Other authors prefer to scrape off the spot and extract the lipid repeatedly with methanolic solvent mixtures*s5 or each area of separated lipid may be directly ashed and the released phosphate determined+*. A previous communication from this laboratory0 gave a preliminary account of both the principles and application of an alternative method providing for the complete recovery of intact lipids and their estimation after prior separation on thin layer chromatograms. In this paper further details of the present improved technique are given together with the analysis of various tissue lipids. Special attention has been given to the separation of certain minor lipid compounds. METHODS

Protein free, washed extracts of human cerebral white matter (internal capsule), human erythrocyte ghosts and rat liver were prepared in chloroform-methanol (z: I, v/v) by the method of FOLCH, LEES AND SLOANE-STANLEY~O andk%BsTER AND I;oLcH~~.

The technique of thin layer chromatography was as previously describedO. Chromatographic glass plates (20 x 20 cm) were layered with 0.5 mm thick Kieselgcl G (Merck). Lipid extract (o.z-0.5 ml) containing not more than 5 mg was placed as a strip (3 cm long) at the origin and the chromatogram developed at room temperature. The various solvent mixtures employed for developing the chromatograms are shown in Table I. After drying the plates, spots or zones were generally made visible by exposure to iodine vapour” except where iodine could interfere in subsequent analysis. For identification purposes, more specific detection methods used were similar to those reported by SRIPSICI el! al.4and SKLDMORE AEZD EN-~ENMAN~~; in addition chromatograms

of pure After

marked x

staining

5 cm),

J, Chvonzalog.,

‘each lipid ~1 lich

lipid

lipids

off. %he plates

containing (3

authentic

with

were was

held

were

iodine,

air dried

transferred up to 50 ml

27 (x967) 385-397

run

as markers. a rectangular

to remove

most

zon.e

enclosing

of the iodine

to a glass

column

(I

of eluting

solvent,

Traces

x

IO cm)

each

and

with

of silicic

spot

the

was

adsorbent a reservoir

acid

and

any

QUANTITATIVE

TABLE TWIN

THIN

LAYER

CHROMATOGRAPHY

OF

LIPIDS

389

I

LAYER

CHROMATOGRAPWY

OF

LIPIDS

Authentic samples of lipid were run on silica gel plates with ascending developing solvents. Mixture A contained chloroform-methanol-ammonia 32.5 y0 w/v in aqueous solution (17: 7: I, by vol.) ; mixture B contained chloroform-methanol-ammonia 20.2 O/( w/v in aqueous solution (17: 7: I, by C contained chloroform-methanol-water (14: G: I, by vol.); solvent D was 1,2vol.) ; mixture dichlorocthane. The temperature of the chromatographic tank was maintained at 18-20~. The identity of each separatecl lipid was checked by use of appropriate spray reagents and was compared with Rp values of authentic samples of known lipids. The mean RF values indicate the order of separation and allow comparison of the three systems. Li$id

Cholesterol Triglyccridc Cholesterol

RF vnlues

ester

Cerebrosides Cardiolipin Ethanolaminc phospholipid Sulphatidc Lecithin Sphingomyelin Ethanolaminc lysophospholipid Phosphatidyl inositol Lysolccithin Serine phospholipid Serine lysophospholipid Phosphntidic acidI

A

B

C

D

0.99 -

-

0.95

0.97

0.95 0.79 0.72 0.84 0.58 0.45 0.33

0.8 0.55 0.20

O.G.1 0.55 0.55 0.45 0.45 0.34 0.18 0.18 0.11

(

0.1 I o.oG

0.03 0

0.70 O.GI 0.50 0.40 0.40 o-33 0.15 0.15 0.15 0.15 0.07

1

r0

0.20 0.2G 0.20 0.15 0.10 -

,

0

0

ELUTION SOLVENT MISTURES Each silicic acid zone containing a single lipid should be treated with solvents as described C = Chloroform ; E = ethanol ; M = methanol ; W = water; N = 32.5 y. aqueous ammonia Lipid

Solvevll covvz~osilion

.Volume (vnl)

Cholesterol Ccrebrosides Cardiolipin Ethanolamine Sulphatide Lecithin

C

20

phospholipid

Sphingomyelin Ethanolamine lysophospholipid Phosphatidyl inositol Lysolccithin Seriue phospholipid Serine lysophospholipid

C-M-W C-M-W C-M-N

(7: 7: J.) ii: ;: Ij

(18:G: I) C-M-W (7: 7: I) C-M (I:I) E-C-W (5 : 2 : 2) C-M (I:I) E-C-W (5 : 2 : 2) E-C-W (5 : 2 : 2) C-M (1:4) E-C-W (5 : 2 : 2) E-C-W (5 : 2 : 2)

below. (w/v).

30 ;: 30 30 5 30 20 20 20 20 15

J. Cltromatog.,

27 (1967)

388-397

390

M. L. CUZNER, A. N. DAVLSON

lipid remaining on the area of exposed glass plate were removed by cleaning with fat free cotton wool damped with chloroform and transferring this pad to the top of the columns. The columns were so packed as to allow slow passage of solvent mixtures, and elution overnight proved to be most convenient (Table II). Phwnalogem For the estimation of individual plasmalogen lipids, plates were not stained with iodine. Lipid samples and markers were spotted on the middle two and outside two lanes respectively of the thin layer plate. After the plate was developed and dried, the middle lanes were covered with a plate of glass or a wide strip of cellophane and the thin layer chromatogram stained with iodine (I 76 w/v in petroleum ether b.p. IOO1~0~). Visible lipid zones were marked off and the iodine allowed to fade. The covering was removed and using the outer lanes as guides, the lipids in the middle lanes of the TLC were marked off. The plasmalogen lipid areas were removed, transferred to columns, eluted and evaporated under nitrogen at room temperature. The plasmalogen content was immediately determined. The methocl was checked by separating human white matter lipids and determining the plasmalogen content of the ethanolamine phospholipid zone. All the phospholipid could be recovered as phosphatidal ethanolamine indicating that no breakdown had occurred; in addition the results agreed with those reported previously 14.Other samples of rat myelin were analysed by quantitative thin layer chromatography and the reliability of the technique again confirmed. Seco9zd stage clzronzatogva$&y 0% alacmina The same type of glass columns were chromatography-BDH, Poole, England) be rechromatographed were transferred to fractionated by stepwise elution, using the

packed with G-IO g of alumina (alumina for and the powderecl silicic acid-lipid zones to the top of the alumina. The lipids were then solvents shown in Table III.

Am2ytical methods Details of analytical methods used are describecl in previous papers. Phospholipid phosphorus was determined by the method of MARTLAND AND ROBINSON~“, cholesterol and its esters by the method of SPERRY AND \VEBI+ and the sphingosine content of cerebroside and sulphatide was estimated using the method of LAUTER AND TRAM+‘. Plasmalogens were determined by the iodine absorption method of WILLIAMS, ANDERSON AND JASIIC~~, primary amino groups by the ninhydrin method of REIODES AND LIXA~Oand ester groups, using the hydroxylamine reagent of SI.IAPIRO~~. Total galactose content of lipid extracts was determined by a method based on those of EDGAI@ and SVENNERHOLM~~ and the choline content of lecithin was determined by the method of SMITS~~. Inositol was released from lipid samples by acid hydrolysis and assayed with KZoecizera brevis, using the method of CAMPLING AND NISON~~. Lipicl samples were hydrolysed in sealed tubes with 4 ml of 50 y0 v/v HCl at 110’ for 6 11.The tops were broken off the sealed tubes, which were subjected to a stream of air in a boiling water bath until the volume was reduced to about I ml, This removed a large proportion of the HCl and the hydrolysate was then completely neutralized with iI? NaO1-L The samples were then added to inositol assay medium 333 (Difco Laboratories) and assayed with K. bvsvis. J. Ckvomalog., 27 (1967) #!8-3g7

QUANTITATIVE TABLE

THJN

LAYER

CNROMATOGRAPHY

OF LLFLDS

39=

III

FRACTIONATION

OF

LIPID

SUBFRACTIONS

ON

ALUMINA

AFTER

THIN

LAYER

CNROMATOGRAPNY

Mixtures of pure lipids were analysed before and after separation on thin layer chromatography. Duplicate samples run on chromatogrsms were transferred as silicic acid areas to the top of prepacked alumina columns. Solvent mixtures used were E = chloroform-methanol (I : I, v/v) ; solvent F = chloroform-methanol (I : 4, v/v) ; solvent G = chloroform-methanol-water (7 : 7 : I, by vol.) ; and solvent I-I = ethanol-chloroform-water (5 : z : 2, by vol.). NO.

I 2 I$_2

3 4 3-1-4 2

Lipid

XolaE Zi$id applied (pmoles)

Li$id recovered from thin layer gZate (/imotes)

Lipid recovered after transferring silicic acid zone io alabmina coZumn and eluting with the respective solvents (Mumoles)

0.004

Pure lysopl~ospl~aticlyl ethanolamine Pure sphingomyelin

0.383 0.830

0.380

Pure lysolecithin Pure phosphatidyl

0.275

0.500 0,270

-

0.240

-

0,625

White White amine

inositol

matter sulphntido matter ethanolphospholipid

0.81

I

E A (zoo ml) (80 mZ)

N (100 ml)

0.380

0.800 0.795

0.382 0,473 0

0.438

0.207

5 -I- 6

Fig. I. Thin layer chromatograms of (I) rat ghost lipids. Developing solvent, solvent aqueous solution (17: 7: I, by vol.). 0 = ; 4 = lecithin inositol ; 3 = sphingomyelin Quantity of lipid applied: 7 = cholesterol.

G (80 ml)

0.274 0.275 0.586

liver, (II) human white matter, and (III) crythrocytc 32.5 o/o w/v in A: chloroform-methanol-ammonia Origin; I = serine phospholipid; 2 = phosphatidyl phospholipid ; C = cerebrosides; ; 5 = ethanolamine 5 mg. J. Ciwomalog.,

27 (x967) 388-397

M. L.

CUZNER,

A.

N.

DAVlSON

Pig. 2. Thin layer chromntograms of (I) rat liver, (II) human white matter, ancl (III) crythrocytc ghost lipids. Dcvcloping solvent, solvent I3 : chloroform-methanol-ammonia 20.2 o/o w/v in aqueous and phosphnsolution (I 7: 7 : I, by vol.). 0 = Origin ; I = scrine phospholipid ; 2 = sphingomyelin ; G = ccrebrosidcs ; phospholipid; 5 = cardiolipin tidy1 inositol; 3 = lecithin ; 4 = ethanolamine 7 = cholesterol. Quantity of lipid appliccl: 5 mg.

Fig. 3. Thin layer chromstograms of (I) human white matter and (11) rat liver. Developing solvent, solvent C: chloroform-methanol-water (14 : G : I, by vol.), I I Phospholipids ; 2 = lecithin ; 3 = and cerebrosides; G = cholesterol. phospholipid ; 5 = cardiolipin sulphatide; 4 = ethanolamine Quantity of lipid applied: 5 mg. J. Cl~vom&og.,

27 (IgG7)

388-397

QUANTITATIVE

THIN

LAYER

CHROMATOGRAPHY

OF LIPIDS

393

RESULTS

In prehminary studies (Table I) mixtures of chloroform and methanol with water or aqueous ammonia in different proportions were compared as developing solvents for individual standard lipids and for mixed phospho and galacto-lipids on Kieselgel G thin layer plates. I,z-Dichloroethane was used as developing solvent for non-polar lipids. Solvent mixtures A and 53 were found to be particularly reliable for the separation of individual phospholipids and cerebrosides and were both employed in the analysis of the lipids of varied tissue fractions, such as myelin, microsomal membrane and erythrocyte ghosts. The second system (B) which resolved cardiolipin at the expense of monophosphoinositide, was usecl for the separation of mitochondrial lipid, which contains cardiolipin as a major component. The cl~loroform-methanolwater system (C) was particularly reliable for the separation of both cerebrosides and sulphatides, but frequently clear separation of phospholipid was not seen. Pwity

md identity of sq?mrated Z@ids Lipids separated by chloroform-methanol-32,s yO w/v aqueous ammonia : 7 : I) were eluted and found to run as single spots in different solvent systems. The (17 individual lipids were identified firstly using specific chromatographic spraysl”. The Rp values of the identified lipids were checked, using pure lipid markers, obtained from many sources. Each separated lipid was analysecl for its chemical constitution. The analysis recorcled in Table IV shows evidence for the identity and purity of separated lipids. Secomdwy se$aration of mixed Zi$ids Where the amount of lipid was only just sufficient for one thin layer plate it was necessary to devise an additional fractionation step to obtain complete separation; this may be done by re-running eluted lipids on suitable thin layer plates or by column chromatography. No~z$oZar @ids. Triglycerides, cholesterol and its esters may be separated by further stepwise elution chromatography on aluminazb, on florisiP or on thin layer plates using carbon tetrachloride or I,z+dichloroethane for development. M&or phos@oZi~ids. In analysis of complex lipicl mixtures, even with optimum solvent systems some overlapping of spots may occur. Three examples of second stage chromatography are shown in Table III. Samples of pure lipid standards (which run together in solvent mixture A) were applied to thin layer plates and after visualization the lipid zone from the chromatogram transferred to a pre-packed alumina column, Resolution of the mixed lipids was then obtained by stepwise elution of the alumina column. Quantitative recovery of mixed Zi$ids 13~11~extracts of human white matter and rat liver were prepared. Repeated analysis of each sample was achieved by the procedure described above (Table V). The method was further tested by mixing known volumes of the analyzed extracts and determining the composition of the resultant mixture (Table V). Lipid extracts, whose compositions are well established, were next prepared and analyzed in order to test the general reliability of the technique. Table VI shows the analysis of such lipid J.

Clwomalog.,

27

(IgG7)

388-397

s

J

w g

z s-

4

2 i 8.’

c

SINGLE LIPIDS SEPARATED

ON KIESELGEL

G THIN LAYER PLATES

Cerebroside

Spot 6 + 7

0

spot 5 Ethanolamine phospholipid I

Lecithin

I

I

spot 2 Phosphatidyl inositol

spot4

I

-.

1-j

2.2

2.3

I.9

Phosphorus Ester

Mokavrafio

Sesine phospholipid

spot I

Separated lipid

-

-

-

-

0.9

Seriae

-

-

O-94

-

-

Choline

-

0.95

-

-

0

Etharrolamirte

-

I.2

-

-

o-9

A ntirle

-

1.04

-

-

-

Plaswalogert

-

-

-

1.2

-

Ittositol

LOS

-

-

-

-

Galaclose

I

-

-

-

-

Sphitlgoske

Lipid extracts of human white matter were separated and individual lipids recovered by elution of silicic acid zones on columns using appropriate solvents.

ANALYSIS OF

TABLEIV

QUANTITATIVE

THIN LAYER

CHROMATOGRAPHY

lipid extracts literature.

in comparison

OF LIPIDS

with the most widely

395

accepted

figures available

in the

DISCUSSION

Methods of quantitative thin layer chromatography generally employed necessitate the destruction of separated lipid whether by charring or by digestion of separated silicic acid zones. Alternatively it is possible to recover intact lipid by repeatedly sh.aking silicic acid zones with solvent, centrifuging and re-extracting until lipid recovery is complete. This procedure is time consuming and moreover losses can easily occur during shaking and transference of eiuting solvent. TABLE

v

QTJANTITATIVE

lipid)

RECOVERY

OF MIXED

Samples of lipid extract or mixed methanol-32.5 o/o w/v ammonia

Lipid

Cholesterol Cerebroside

Total phospholipicl Serinc phospholipicl Phosphatidyl inositol Lysolecithin Sphingomyelin Lysophosphatidyl cthanolaminc Lecithin Ethanolamine phospholipid Cardiolipin

LIPID

EXTRACTS

FROM

TI-IIN

LAYER

CHROMATOGRAM

extract Were run on thin layer chromatogrsms using (I 7 : 7 : I, by vol.). Analytical methods are as described

Human white matter (3-6 ~92s)

Rat liver

o&87 0.305 0,508 0,092

0.151 -

& & & Ifi

0.025 0.033 0.052 0.002

0,010

0.004 0.089 0.020 0,122

& 0.004

0,167

0.013 -& 0.015

0,012 x01.5

o/0 recovery

&

(3-5

0.035 I02

chloroformin the text.

Mixtzcve of humnn white matter and rat Eive~ (2-3 runs)

~vuns)

0.833 0.031 0.052 0.008 0.022 o.or.6 0.481 0.206

(~n’lOleS/ITlg

f f f

0.022 0.006 0.007

f

0.001

-+ 0.018 rt: 0.012

0.790 0.118 0.061 0.010 0.116

0.037 0.570 0.360 0.04G

o/o recovery

The methods described in this paper are designed to provide a rapid and reproducible means of quantitative separation of complex lipid mixtures. Recovery of lipids from silicic acid zones by elution in columns has four advantages: (I) it is time saving, in avoiding repeated shaking and extraction of silicic acid; (2) it results in the recovery of intact lipid and hence does not limit the use of analytical methods; (3) silicic acid is eliminated from the recovered lipid and’hence it cannot interfere with colour development, or add to blank values; and (4) recovery of intact silicic acid-free lipid enables the method to be used for separation of labelled lipid. Previously completeness of recovery of phospholipid has been checked with 32P labelled lipid”. In this paper fresh data gives detailed evidence of recovery based on repetitive chemical analyses. The use of columns has also been adopted for a further subfraction akin to a diethylaminoethyl two-dimensional chromatogram. ROUSER et al. 36 have employed (DEAE) cellulose column chromatography as a primary step followed by quantitative thin layer chromatography. Spots produced are charred with sulphuric acid-potassium dichromate and measured by quantitative densitometry. The method described in this paper utilises the principle of column separation after thin layer chromatography. J, Cltvomatog.,

27 (x967)

388-397

molau

0 2.2

0

2s

3-t-7

20.1

‘4-3 I.4

100

89’

(I

(2 YUllS)

-

44.’ 31-5 54

5-3 7.5

100

11.1

run)

Pvesea! study

Literatfave27-30

r-7

24.8

89.5 100 II.2 4.7 27.2 32.5

Preseflt sllrdy (3-6 YWZS)

-

2.7 6.0 4.6 51.6 30-4 5.4

100

II

(4 YlWS)

Lifeeuatur&3”

Rat liver witochoj~dria

rafios (phospholipid = xoo)

Hmatr erthrocyte.ghosts

Lipid

* Mean figures calculated from VANDEENENAND DE GIER~;.

Cholesterol Cerebroside Sulphatide Total phospholipid Se&e phospholipid Phosphatidyl inositol Sphingomyelin Lecithin Ethanolamine phospholipid Cardiolipin Phosphatidic acid

Lifiid

49

-

3.1 15.4 19.3 35.5

IS3

LOO

-

10s

(3 r1m)

Pvesewtstudy

-

-

43

24

I55

2.0

14

13G 51 7 100

i#-6

YlM)

Literahue33~x

3’

LOO

48

1

rro=

(5 YUfZS)

The tissue extracts were prepared and lipids analysed by the method given in the text. Total phospholipid has been equated to a figure of LOOmoles % and the molar ratios of all other lipids calculated on this basis.

ANALYSIS OF VARIOUS TISSUE LIPIDS

TABLE VI

QUANTITATIVE

TI-IIN

LAYER

CWROMATOGRAPWY

397

OF LIPIDS

This second stage separation has been successfully applied to resolution and analysis of lysophosphatides from other phospholipids, similarly it is possible to separate cholesterol, cholesterol esters and triglycerides on columns or by further running on thin layer plates. ACICNOWLEDGEMEi\iT

The support.

authors

wish to thank

the

Medical

Research

-Council

for their

generous

SUMMARY

A method for the quantitative separation of tissue lipids is described. The technique is based on preliminary separation of total lipids on thin layer plates of silicic acid with, when necessary, a second stage column chromatography. This simple procedure allows the complete recovery of intact single lipids from naturally occurring mixtures. REFlmENCES

I 0. S. PRIVETT, M. L. BLANK, D. W. CODDING AND E. C. NICK~LL,J. Am. 42 (19’%) 381. 2 J. CLAUSEN, I-I. 0. C. Lou AND W. ANDERSEN, J. Nczwochem., 12 (1965) 3 A. N. SIAI~OTOS AND G. ROUSER, Anal. Riochem., 14 (19%) 162.

4 5 G 7 8 g

Oh? Cltcmists’

Sot.,

599,

V. P. SKIPSKI, R. I;.PETERSON AND M. BARCUY, ~iocltem.J..go (1964) 374. L. SEMINAIZIO,E. I?.SOTO AND T. DIZ COHAN,J. Cltromalog., 17 (1965) 513. I-I. JATZKEWITZ,%. Physiol.Chem., 326 (IgGI) GI, M. J. EVANS AND J. 13.FINEAN, J. Newockem., 12 (1965)720. W.T. NORTON AND L. A. AUTILIO,J. Ncuvochem., 13 (1966) 4. A. N. DAVISON AND E.GRAFIAM-WOLPAARD, J.Ncuvochcnz., II (1964) 147,

IO J. FOLCEI, M. LEES

AND G. 1-I. SLOANIZ-STANLEY,

J, Biol.

Chem.,

22G

(1957)

$97.

G. R. WIZBSTIZR AND J. FOLCH, B,iocltim. Biopliys. A&a, 4g (1961) 3gg. G. I-I. DEHAAS,I?.P.M.BONSEN AND L.L.M,VANDEIZNEN, Biochim.Bio~ilys.Acla, I 1G(1g66) 114. 13 W. D. SICIDMORIZAND C. ENTENMAN,J, Lipid RES.,3 (1962) 471, 14 G. R. WEBSTIZR, Bioclha. l3iop7tys. ,4cla,44 (1900) rag. 15 M. MARTLAND AND R. ROBINSON, Bioclrem. J., 20 (1926) 847. 16 W.M. SPERRY AND M. WEBB,J. 13ioZ. Chem., 187 (1950)97. 17 C. J. LAUTER ANT) E. G-TRAMS, J.Lipid RES,, 3 (IgG2)136. 18 J. N. WILLIAMS, C. E. ANDERSON ANKI A. D. J,\SIIC, J, L,ipid Rcs., 3 (1962) 378. xg D. N. RHODES AND C. I-I. LEA, Biocltem. J.. 56 (1954) 613. 20 B. SWAPIRO, Biochem. J., 53 (1953) 663. 21 G. W. I?.EDGAR.AC~U Anai.. 27 (1956) 240. 22 L. SVENNERHOLM, Nahcre. 177 (1956)524. 23 G. SMITS, Biocltinz. BiopJkys. Acta., 26 (1957)424. 24 J, D. CAMPLING AND D, A. NIXON,J. Pl&ysioZ. (Londou), 12G (1954)71. 25 C. W. M. ADAMS AND A, N. DAVISON,J. Netwochem., 4 (1959) 282. 26 II;. IC.CARROLL, J. Lipid Res.; 2 (I~GI) 135. 27 L, L.M.vAN DIZENEN AND J, DE GIER, inc. BISI-IOP AND D.M. SURGENOR (Editors),The Red Blood Cell, Acaclemic Press,New York, 1964, p. 243. 25 G. B. ANSELL AND J, N. HAITI-IORNE, Pltos$holipids, Elscvier,Amsterdam, 1964, p. 210. 2g R.M.C. DA~soN,N. HEMINGTON AND D. 13.LINDSAY, Biocltenz.J., 77 (IgGo) 226. 30 R.M. C. DAWSON,N. HEMIMGTON AND J.B.DAVE;NPORT, Biochem.J., 84 (rgG2)497. 31 L. A. E. ASHWORTH AND C. G~~~~,Scicncc, 151 (sgGG)210. 32 M. I. GURR, C. PROTTEYAND J. N. HAWTHORNE, Biochim. Bio#7&ys. Acta, IOG (rgG5) 357. 33 W. T. NORTON AND L. A. AUTILIO, J. Nearroc7wa., 13 (1966) 213. 34 E. I?.SOTO, L. SXMINARIO AND M. CALVINO, J. Nez#ocltem.,13 (1964)gSg. 35 G. ROUSER, C. GALLI, E.LIIx~ER,M. L.BLANKAND O.S.I?RIVETT, J.Am. Oil C7tcmisis’Soc., 41 II 12

(1964)

836.

J. Chromalog.,

27 (IgG7)

388-397