Colorimetric Determination of Glucosamine BENJAMIN SCHLOSS' New York University College of Medicine, New York, A'. Y .
H E first practical method for the determination of glucosTarnine was the colorimetric method developed by Elson and Morgan (5) i n 1933. Since then several modifications have been described (3, 8, 9, 11, 11). These methods are superior to other types (1, 4 , 7 , IS), mainly in sensitivity and simplicity. Boyer and F k t h ( 3 )following the Elson and Morgan procedure were unable to determine glucosamine in the presence of hydrolyzed protein. Nilsson (@, Palmer et al. (9), SOrensen ( I I ) , and Blix (8) overcame this limitation with their more sensitive modifications. They failed to note the complexity of the reactions in this method, with the resultant enhanced possibilities of error, although they reported serious discrepancies in the analyses of biological materials. The reliability of their methods is questionable. The author has investigated the reactions used in the Elson and hlorgan method with the aim of developing a precise, sensitive method that can reliably determine hexosamine in complex biological materials. Aside from several procedural inodifications, the method presented here is similar to those developed by earlier workers. An important addition is the criteria for testing the reliability of the method. Perhaps of greater significance is the discovery of the formation of a volatile chromogen in the Elson and Morgan reaction. This pioperty probably can be used to develop a highly specific method for glucosamine, that may possibly serve to distinguish between glucosamine and galactosamine.
use of impractically large volumes of reaction mixture), two nonchromogenic white crystalline solids, a white crystalline chromogenic solid, an amorphous dark brown chromogenic solid, and a volatile chromogenic liquid, and found evidence of the presence of another chromogenic solid. ISOLATION EXPERIMEYTS
Two hundred milliliters of aqueous solution containing 5.4 grams of glucosamine hydrochloride were added to 550 ml. of acetylacetone reagent prepared from 26.5 grams of sodium carbonate, 4.9 ml. of acetylacetone, 50 ml. of 1.0 N hydrochloric acid, and water. The p H was 9.75. The mixture was refluxed in a boiling water bath for 20 minutes, chilled, and distilled under pressure. The liquid distilled under 30 O C. The first 200 ml. were collected in a flask immersed in dry ice-ethyl alcohol mixture. The next 400 ml. of distillate were discarded (tested negative for chromogen). The residue (Mixture I ) gave a negative pine shaving test. The distillate was extracted with five 20-ml. portions of absolute ether, the water residue was discarded, and 5 grams of calcium chloride were added to the ether extract and stored overnight at 0' C. The solution was filtered and the calcium chloride residue discarded. The solution was distilled under reduced pressure a t less than 0" C. into a flask immersed in dry ice-ethyl alcohol mixture. Toward the end of the distillation, the flask was immersed in a water bath a t 40" C. The distillate was discarded. The residue was approximately 0.3 ml. of a viscous, pale yellow, unstable intensely chromogenic liquid (Liquid I). It turned pale brown after standing for 24 hours a t 0 " C., and on heating, it rapidly changed into a dark brown tar. No boiling point was observed up to 100" C. a t a pressure of 30 mm. of mercury. Considerable decomposition occurred. The liquid had a strong, peculiar odor. I t gave positive pine shaving, Ehrlich's, and phenanthraquinone tests and negative results with fuchsin and p-dinitrophenylhydrazine. Approximately 0.2 ml. of Liquid I vielded 80 mg. of a twice recrystallized (from hot water) phthalide melting a t 205-206" C.
NATURE OF T H E REACTIONS
The Elson and Morgan method and it$ various modifications are performed by first heating a glucosamine hydrochloride solution with an alkaline solution of acetylacetone (Reaction I) and then adding an alcoholic-acid solution of p-dimethylaminobenzaldehyde (Reaction 11). A chromogen is formed in Reaction I and a red solution is obtained after addition of the aldehyde. The color intensity is measured after a period of incubation. Elson and Morgan thought that the red solution resulted from the condensation of 3-acetyl-2-methyl-5-tetrahydroxybutylpyrrole with p-dimethylaminobenzaldehyde, because Pauli and Ludwig (10) claimed it was obtained from glucosamine heated with an alcoholic solution of acetylacetone. They did not note that the conditions in their reaction mixture were appreciably different from those used by Pauli and Ludwig. Boyer and Furth (5) later showed that Pauli and Ludwig had performed their analysis on an impure compound and had used the wrong empirical formula for the pyrrole. It can easily be shown that more than one chromogen is formed. When the first part of the Elson and Morgan procedure is followed and the resulting solution is distilled under reduced pressure, a volatile chromogen distills over in the first fractions of the distillate and a nonvolatile chromogen remains in the residue. With p-dimethylaminobenzaldehyde solution, the distillate forms a purplish red solution absorbing maximally a t 550 mp, whereas the residue forms an orange-red solution absorbing maximally a t 512 mp. The absorption curves of the colored solutions from Reaction I1 were found to change in a complicated manner as the p H and period of heating of the Reaction I mixture were altered, indicating greater complexity in the chromogen-forming reaction. The author isolated from Reaction I mixtures very similar to those used in the analytical procedure described below (glucosamine hydrochloride concentration was increased 1000-fold to avoid the 1
The absorption of the colored solution obtained with Ehrlich's reagent mas maximum a t 550 mp. The color developed slowly but was stable after maximum intensity was reached. The absorption curve was not quite symmetrical, the absorption decreasing more sharply toward the red than toward the violet. An important property of this colored solution was the reversible shift in the wave length of maximum absorption with varying acid concentration. The solution changed from blue violet in slightly acid solution through purplish red to yellowish pink in concentrated acid. The colored solid (Solid I ) was isolated by reacting the distillate from Reaction I with p-dimethylaminobenzaldehyde in aqueous hydrochloric acid solution (final acid concentration, 4.4N ) , incubating for 3 hours a t 30" C., neutralizing with sodium carbonate, and filtering, The brown residue was dissolved in chloroform and reprecipitated with ether. This was repeated until a constant nitrogen content and specific extinction coefficient were obtained. The per cent nitrogen was 11.0 and the specific extinction coefficient a t 550 mp was 33,400.
Present address, The Nucleonic Corp. of America, Brooklyn 31, N. Y.
Solid I decomposed without melting. It was insoluble in water, ether, acetone, and benzene; very slightly soluble in methanol and isopropyl alcohol; somewhat soluble in hot ethyl alcohol, pyridine, and morpholine; and very soluble in chloroform and dilute hydrochloric acid. Attempts to obtain this solid in crystalline form were unsuccessful. From the residual liquid (Mixture I), 80 mg. of dark brown clumps of solid (Solid 11) were mechanically removed. Solid I1 was intensely chromogenic. It melted a t 145' C. but appeared to decompose first a t 90" C. With Ehrlich's reagent it gave a colored solution that absorbed maximally a t 530 mp. The formation of Solid I1 appeared to be a function of the concentration of
1321
ANALYTICAL CHEMISTRY
1322 glucosamine; its yield seemed to vanish as analytical concentrations of glucosamine were approached. After the clump? of Solid I1 had bern removed, the remaining liquid was distilled under reduced pressure. The moist, light yellow residue was dried for 24 hours at rrduced pressure ovrr phosphorus pentoxide. The dry rwidue \vas extracted five timrs wit,h 50-ml. portions of absolute ('thy1 alcohol, 750 ml. of absolute ether were added to t h r estrart, and thc supernat'ant' (Supernatant, I ) was decanted. The precipitate was stirred wit,h 50 ml. of absolute alcohol and filtered. The inorganic residue was discarded. The filt,rate, evaporated a t room temperature with a stream of dry air, yielded crystnlr: in a brown oil. After 10 ml. of alcohol were added and the mixture was filtrred, 0.7 gram of yrllow crystals melting a t 55-5G" C. (Solid 111)was obtained. The filtrate was chromogenic, but, attempts t,o isolate crystalline material were unsuccessful. Solid I11 n-as obt'ained as white crystals t)y redissolving in 30 nil. of absolute ethvl alcohol, stirring with 0.1 gram of active carbon (Darco GGO),and filtering. Evaporation of the filtrate yielded nonchromogenic white crystals melting at 55-56" C. Solid 111 gave a yellox crystalline picrate, which aftrr thrrr rrcrystallizatione froni alcohol melted a t 161-162" C.
*
.09
E
N
In
-x Em
.04
z
(after Solid I1 had been removed and the remaining water distilled off) with p-dimethylaminobenzaldehyde in aqueous hydrochloric acid. The solution was incubated at 30" C. for 24 hours. The remaining procedure was 5iniil:ir to that used for isolating Solid I . The per cent nitrogen \vas 7.12 and the specific extinction coefficient at 512 mp (masimuni nbsorption) was 8910. .ikhough the aut,hor Puc~c~c~cttied in isolatiiig only two chromogenic solids from l l i s t u i ~ cI,~thew wis colorimrtric evidenre that a third importaiit chi~omiigeiimight tie prcsrnt. After Renction I had bern performed with :i yluco~amiriesolution of 40 micrograms per ml. and a l l the liquid h:id lieen clistilled off, the residue rapidly formed a colorrd solut,ioii \\.Iic.ti :iridic p-diiiiethyl:tmiiiol~enz:tld~:hyde solution w a ~ :idded. Th(. :il)sorption curve was symmetrical and was :it :t ni:iximuni :it 530 nip. When this colored solution W R S inrul)atetl for 24 hour.^. :i iii:i\- ~yininrtricalabsorption curv(~\ v m present Ivitli :L ni:txiniuiii :it 312 nip. The first atxwrption CUI'VC had di~:ippr:iri~(l. 1~1~11iii t l i c w ,)i)srrvetions one can concluck either t,ti:it t w i is111 otiiogriir : I W 111 nt, or that from one chromogen, n colorid c-ornpouii(1 .I i, f( ell xhich tlec:tyP to f o r m a colored conipourid 1%. It is iiioi'i, likely that two chromogens are present. Chnipoiuid .I:ip1i~:ii,rdto decay f:ister than compound B grew. Fui~theriiii!rt~,S$t~cnsoii ( 2 1 ) found that the gre:itcst protluc.rti when the Reaction I niixtme yield of compoulid .I was a t pII 9.5, wtierca for the formation of ro The results of tlie :thivc>cspcrinients can be summarized as follows: In Rcart,ion 1arc fornietl a chromogenic liquid (Liquid I), two nonchromogenic solids (Solids 111 arid IV), two chromogenic solids (Solids I1 antl VI), :tiid prohalily a t,hird chroinogrnir solid (not isolated).
w
0
ANALYlICAL AIETHOD
-I
So complex a reaction \vould appear to oyer litt.le promise for a precise analytical method. However, a sufficient latitude in the significant variables was found PO that reproducible 1-ields of chromogens and colored compounds could he ohtainetl.
a
0
ca .o: 0
.02 I 75
I
I
I
I
80
85
90
95
I
100
P" Figure 1.
.
Color Intensity with \ a r y i n g pll of Reaction I Miutiire
Supernatant I was distilled under reduced pressure into a flask immersed in a dry ice-ethyl alcohol mixture. The nonchromogenic distillate was discarded. The. sirupy residue u-as diluted with 10 ml. of butanol. About 10 minutes after adding 5 ml. of methanol, white crystals separated out. Filtration yielded 0.2 gram of nonchromogenic white crystals (Solid I V ) melting from 145" t o 158" C . Recrystallization with butanol and methanol yielded transparent white cryqtals melting from 148" to 150" C. Then 50 ml. of absolute ether were added to the filtrate, the supernahnt liquid was decanted and discarded, and the precipitate was dissolved in 10 ml. of butanol, evaporated to 5 ml., and filtered; 80 mg. of white cryptals melting from 144"to 146' C . remained in the residue. To the filtrate 50 ml. of absolute ether were added. The supernatant n-as discarded. The hygroscopic precipitate rapidly reverted to a bronn sirup. Then 20 ml. of methanol were added antl the solution was stirrrd three times with 0 1-gram portions of active carbon (Darco (260). To the resultant pale yellow liquid werc added 150 ml. of ether, 1.5 grams of a light yellow, inteniely chiornogenic precipitate (Solid V ) were obtained. Attempts to isolate crystalline niatrrial from this solid were unsuccessful. However, Rhcn 50 ml. of absolute ether were addrd to the supernatant and the mixture u a s stored overnight a t room temperature, 3 mg. of a crystalline, white, intensely chromogenic solid (Solid VI) were formrd Solid VI melted from 90" to 95' C. It was very hygroscopic. The chromogenicity of Solid V $1 a s rapidly destroyed by concentrated hydrochloric acid. .4 colored solid (Solid TIT) n :I. iwlatrd hy rraction of Xlixturr I
Reagents. Standard glucosaminc~hydrocliloride solutions \yere prepared from sis times ~ccrystallizcdEastman Iiodak glucosamine hydrochloridr. C;raviinc,t i,ic, Iijrldahl, formal titi,ation, and specific rotation &it erniiriatioiii: of the final recryst allized present were less than O.Ic;. product indicated that inipur Reagent -4 wis prrp:~rcd1)). olving uwtylac.rtonr in 1.0 S sodium carbonate solution. gmt B \van prepared hy dissolving 0.80 grim of p-dinieth~l:iniiiioIii~iiz~tl~leliyd~~ in 30.0 1111. of commercial atmolutr :il(whol :iiid 30.0 i d . of c~)iic.riitr:~trd hydrochloric acid. Palmer et al. (9) h:id noted that acetaldehyde in alcohol interferes with the deternliiintion, decreasing t,he intensity of color. The author found that alcohol containing as much as 30 micrograms of acetaldehyde per nil. of alcohol produced no observable effect. This concentration of aldehyde is considerahly grrater than that found in coninierc.i:il :thsolute alcohol. CHROhIOGEN-FORMI~(: REACTIOS (REACTION I )
I n the experiments described l~elow,absorption mensurements were generally made :it 512 nip rather than the wave length used iiy earlier workers (530 nip) because the compound absorbing maximally a t 530 nip i p not stable and absorption is greater a t 512 nip. This results in optiniuni operating conditions t,hat are different from those previous+ employed. The p H of the Reaction I misture and the period of heating were found to affect thr naturr and intensity of the color formed after the :idditioIi of ltc~ngentH. The color changes that were noted vhen the pII \ W P altered \vrw attributed solely to changes in the yields of chroniogenic m l i s t m r e s , hecause lnrge variat,ions of the acid conccntr:rtion in the Reaction I1 mixture had no appreciahle effect o i i the alisorption curvc (see Tihle I ) . The optical densities were detcrniined in colored solutions that were formed hy the procctlure descriticd I)clo\~,escept that the pH of
V O L U M E 23, NO. 9, S E P T E M B E R 1 9 5 1 the Reaction I misture was changed by varying the pH of Reagent -\as follo\vs: hcetylacetone (1 ml.) was dissolved iii 50 mi. of 1.0 K sodium carbonate solution. Portions of this solution were then added to equal volumes of hydrochloric acid solut,ions of varying normality. The p H of the reagents were measured with a Cambridge pH meter, using a Beckman high alkalinity glass electrode. The glucosamine hydrochloride solution contained 10.0 micrograms of glucosamine per ml.
The nature and the yield of the chromogenic substances appear to he altered as the pH of the Reaction I niisture is varied, as is evidenced by shifting of the absorption intaximum and variation in the int,ensities of absorption a t these maxima. Stable maxima
appear at 540 nik a t pH 7.7 arid at, 512 nip 3t pH 10, whereas a transitory mssinium appears :it 530 nip n t all pH’s. Figure 1 shows the absorption a t 512 nip at various pH’s. There is a small region (pH 9.63 t o 9.98) i n which the intensit,y of absorption is independent of t,he pH. SZrcinsen ( 2 1 ) failed to find such a region using 530 nip as her reference wave length. Period of Heating. When Reagent B was added to Reaction I niistures previously heated for different lrngths of time, colored solutions were obt.airied n i t h v:irying int!ensities of absorption a t 512 nip. hlthough the intensity o f almrption a t 512 nip was the same altw 15 to 30 minuttls ( J f previous heat,ing, the color did c8hiingc~. This became apparent t,o thts rxye as a shift in color from piilk ivitfi 15 minutes’ heating towrrtl :tn increasing purple intensity as the heating \vas prolonged. .Iri examination of the absorption curves showed that the alisoiytion maximum a t 512 mp vas unaltered but absorption had increased i n the 540 mp region and decreased in the blue-violet region. Thus, with t,he two significant variables, pH and the tinic of heating, there are regions in which the yield of chromogen appears t o 1 ) constant. ~
1323 the color. Large variations in the acid concentration (see Table I) had no appreciable effect. Incubation Time. When measurements were made a t 530 mp, within a few minutes after adding Reagent B, the absorption of the colored solution was already close to its maximum value, but soon began to decrease. At 512 mp, however, absorption was low a t first, probably being almost entirely due to the presence of this 530 mp compound. -4hsorption slowly increased until, after about 20 hours, it reached a higher and more stable maximum value. Practically no change in absorption occurred in the following 12 hours.
.08
.061
>
L v, z w
n
,041
-1
a 0 I-
n 0
,021
COLOR-FORMING REACTIOh (REACTION 11)
The alcohol concentration, incubation time, and incubation teniperature were found to affect either thc intensity or nature of
590
550 WAVE LENGTH
Figure 3.
510
cm
470
p )
Absorption Curves at Different Incubation Temperatures
Alcohol Concentration. The effect of alcohol concentration 011 the intensity of absorption is shown in Figure 2. The data were obtained by following the procedure given below, using varying amounts of alcohol. Glucosamine concentration was 40 micrograms per ml. I t is apparent that alcohol has a marked effect on the intensity of the color. Except for the differences in the color intensities of the solutions, the absorption curves in 5 and 67% alcohol are identical. By working with alcohol concentrations above 60 volume %, colored solutions are obtained whose intensities are maximum and are practically independent of small variations in alcohol concentration. Incubation Temperature. I t can be seen from Figure 3 that the color intensity varies with the incubation temperature. Accurate control of this temperature is therefore essential. Maximum absorptions a t 512 nip Tere obtained after only 2 hours’ incubation a t 50’ C. However, the colors were unstable and the absorption intensity was very sensitive to incubation temperature. The above observations were utilized for the development of a precise, sensitive colorimetric method.
I-
n
0
.lo(
PROCEDURE
ALCOHOL
Figure 2.
I
I
I
20
40
60
CONCENTRATION (VOLS. X )
Ahsorption at Varying Concentrations
Alcohol
Deliver 2 ml. of the glucosamine solution (concentration range
4 to 40 micrograms per ml.) from a volumetric pipet into a 25-ml.
borosilicate glass volumetric flask and add from a buret 5.50 ml. of Reagent A containing 0.049 ml. of acetylacetone, 5.00 me. of sodium carbonate, and 0.75 me. of hydrochloric acid. This acetylacetone reagent may be stored in the refrigerator for several
ANALYTICAL CHEMISTRY
1324 days without change. It should be brought to room temperature before using. The acetylacetone fraction boiling a t 138" C. (uncorrected) is used for preparing the reagent. It should be perfectly colorless. If any free hydrochloric acid is present in the glucosamine hydrochloride solution, its concentration must be determined. Less hydrochloric acid is added in preparing the reagent, so that the total amount of hydrochloric acid is still 0.75 me. The p H of the acetylacetone reaction mixture should be 9.8 when measured with a Beckman high alkalinity glass electrode standardized against a p H 9.180 borate buffer. Place the flask in a boiling water bath, making sure that the water in the bath is above the solution level in the flask. The water bath is constructed from a long shallow tray with notches cut out a t the sides. The volumetric flasks are tilted so that their necks rest in the notches and protrude beyond the sides of the tray out of the path of the rising steam. Place a weight on the flask to prevent tipping. Insert a ground-glass stopper after approximately 0.5 minute and leave the flask in the boiling water bath for 20 minutes. Immerse in cold running water for 1 to 2 minutes. These solutions will stand for more than 0.5 hour nithout noticeable change. This is sufficient time for performing a large number of simultaneous determinations. Add absolute alcohol to within several milliliters of the neck of the flask and shake. Add from a buret 2.50 ml. of Reagent B. This reagent can be stored in the refrigerator for several months without noticeable change. I t is allowed to come to room temperature before using. The mouth of the bottle should be carefully wiped clean before pouring, as a purple substance accumulates during storage. Shake the flask carefully a t first to prevent escape of the foaming solution, then more vigorously until the solution is well mixed. Add absolute alcohol to the mark. Stopper tightly and invert two or three times, then add absolute alcohol to the mark if there has been any further change in volume. Incubate a t 30' C. Make absorption measurements 24 hours after adding Reagent B, using a blank prepared by the same procedure but nith water substituted for the glucosamine solution. An ordinary photoelectric colorimeter can be used in place of a Beckman spectrophotometer if one uses a narrow band interference filter with maximum transmittance a t 512 ms.
t.O(
/
a
E (u
1.50
u) I
x
t
k v)
n
1.00
-1
4 .
0
I-
n
0
.50
0
I .o
2.0
c (7) =
D/0.0680 X dilution factor
CHARACTERISTICS OF METHOD
Calibration Curve. The curve showing the variation of the optical densities of the colored solutions with glucosamine concentration is shown in Figure 4. S o appreciable alteration of the curve was found when it was redetermined on several different days. -4straight line (espocted from the Beer-Lambert law) Tas obtained only up to equivalent glucosamine concentrations of 3 micrograms per ml. The concentrations of the glucosamine solutions are here divided by 12.5 to indicate the concent,rations in the final colored solut~ions.
Table I.
Effect of .4cid Concentration on Intensity of Absorption at 512 mp Hydrochloric Acid in 25 M1. of Final Colored Solution. Me.
4.0 9.0 11.5
Extinction Coefficient 0.0561 0.0571 0.0565
Specificity. The specificity of the method w m investigated with respect to some commonly encountered substances. Solutions of urea, glutamic acid, proline, tryptophan, arginine hydrochloride, glycine, lysine, histidine, serine, phenylalanine, leucine, alanine, tyrosine, hydroxyproline, valine, and norleucine (concentration 100 micrograms per ml.) failed to produce any color, nor did they interfere wit,h the determination of glucosamine. S-Acetyl glucosamine gave a color intensity about one fifth that derived from glucosamine with a maximum absorption a t 530 mg instead of a t 512 mp. S-Acetyl glucosamine can also be easily distinguished from glucosamine by observing the effect of heating with acid whereby .\--acetyl glucosamine is converted to glucosamine. The red solution obtained with pyrrole has an absorption which differs markedly in its characteristics from that obtained with glucosamine. Maximum absorption is a t 550 mp. The color fades rapidly and the maximum absorption shifts with time to longer nave lengths. Pyrrole can also be readily distinguished from glucosamine by the fact that it forms a colored solution without initial t.reat,mentwith acetylacetone reagent. Precision and Range. The average deviation from the mean of four replicates was i ~ O . 5 when 7 ~ solutions ranging in concentration from 12 to 830 micrograms of glucosamine per ml. were analyzed. A similar precision was found when twenty further duplicat,e analyses were performed on one of these solutions (20.8 micrograms of glucosamine per ml.). Sensitivity. Glucosamine solutions containing less t'han 4 micrograms of glucosamine per ml. fail to give a readily identifiable absorption curve.
3.0
GLUCOSAMINE CONCENTRATION (NG./ML.)
Figure 4.
sq. cm. per microgram. The concentration in the unknown can then be calculated as follows:
Glucosamine Calibration Curve
The glucosamine concentration in the initial solution is obtained from the observed optical density, D, by first referring to the calibration curve (Figure 4). (This curve should be independently determined by the analyst.) This value is then multiplied by 12.5 and by any other dilution factor to give the concentration in the original unknown solution. Up to D = 0.177 (66.5% 7') the specific extinction coefficient, k , is constant and equal to 0.0580
DISCUSSION
The method presented in this paper differs from earlier methods that employed the Elson and Morgan reactions, mainly in that measurements are made on the stable colored compound absorbing maximally a t 512 mp, whereas all other workers measured the unstable compound absorbing maximally a t 530 mp, and the p H of the Reaction I mixture is adjusted to 9.8, which is in a region where small variations of pH are without effect. Earlier workers generally overlooked the importance of the pH on the reliability and precision of the method. As measurements were made a t 530 mg, small variations in pH were critical (f 1 ) ; hence
V O L U M E 2 3 , N O . 9, S E P T E M B E R 1 9 5 1 even when attempts were made to control the pH, a high degree of precision was not possible. The reliability of colorimetric methods is usually suspect because of the many possibilities of interference. If more than one property can be measured the reliability is increased. The following properties can be used for a reliabilitv check: formation of a volatile and nonvolatile chromogen, shift of the absorption maximum from 540 to 512 mp, with increasing period of incubation, and shift in the yield of chromogen with pH. The author believes that the method cannot be safely applied to the analysis of any unknown without this type of reliability check. Inimers and Vasseur ( 6 ) recently noted serious interference when a sample rontained both glucose and lysine or glucose and vaiious amines. There are serious discrepancies between the glucosamine values of normal human blood serum reported by various investigators. Thp author has applied this method to human blood plasma (paper in preparation). By testing for the properties described above, a more reliable hexosamine value was obtained and the reasons for the reported discrepancies became apparent. One of the chromogens isolated from the Reaction I mixture may possibly be formed only from glucosamine. This could serve to distinguish glucosamine from galactosamine, which the Elson and Morgan method fails to do. A highly specific method for hexosamine could be developed by basing the analysis on the detwmination of the volatile chromogen (Liquid I).
1325 ACKNOWLEDGMENT
The author is indebted to Isidor Greenwald for his helpful suggestions during the invrstigation and his critical review of t h r manuscript. LITERATURE CITED
Bierry. H., and Magnan, C., Compt. rend.
soc. biol., 114, 257
(1933).
Blix, G., Acta Chcni. Scand., 2, 467 (1948). Royer, K., and Furth, O., Biochem. Z., 282, 249 (1935). Dumazert, C., and Lehr, H., T m c . soc. c h i m . b i d , 24, 1044 (1942).
Elson, L. d.,and Morgan, TV. T. ,J., B i o c h ~ mJ. . , 27, 1824 (1933). Imrners, J . , and Vasseur, E., .Tuture, 165, 898 (1950). Kawabe, K., J . Biochem. ( J n p a n ) . 19, 319 (1934). Kilsson, I., Biochem. Z., 285, 3S6 (1936). Palmer, J . TT-,, Smyth, E. >I,, and Meyer, E., J . B i d . Cheni., 119,491 (1937).
Pauli, H., arid Ludwig. E.. Z.phjjsiol. C‘liem., 121, 176 (1922). S$rensen, lf.,Conipt. v e n d . frnz’. lab. Carlsberg, 22, 487 (1938). West, R., and Clai,ke,D. H., J . Clin.Inaest., 17, 173 (1938). Zuckerkandl, F . , and >lessinei-Iilehermass. I,., Biochmn. Z., 236, 19 (1931). RECEIVED August 19, 1950. Preaented before the Division of Biological Chemistry at the 114th JIeeting of the AarsaIcax C H E M I C A L SOCIETY, Washington, D . C. Froin a diusertation submitted to the Department of Biological Chemistry, Graduate School of Arts and Sciences, New York University, in partial fulfillment of the requirements for the degree of doctor of philosoyhg.
Application of lead Reductor to Indirect Determination of Sodium W A L L A C E M. M c N A B B , J. FRED H A Z E L , AND H O R A C E F. D A N T R O University of Pennsylvania, Philadelphia, Pa. OLTHOFF and Sandell ( 2 ) have described a method for the titration of uranium and its application to the volumetric determination of sodium. The sodium was precipitated as the sodium zinc uranyl acetate, filtei,ed, nashed, and dissolved in sulfuric arid, and the sexivalent uranium was reduced to a mixture of tri- and quadrivalent uranium in a Jonep reductor. After aeration to o x i d i z e trivalent to the quadrivalent s t a t e , t h e s o l u t i o n was titrated with a standard potassium dichromate 801 u t i o n , using ferric iron as a “catalyet” a n d d i phen y I a m i n e sulf o n a t e a s a n indicator. As the ratio of uranium to sodium in t h e precipitate is constant, the uranium value was used to calculate the amount of sodium. R e c e n t l y Cooke, Hazel, and McKabb (1) p u b l i s h e d a method for the determination of uranium, using a lead reductor in place of a zinc reFigure 1. L e a d R e d u c t o r ductor. The aeraA . 250-ml. suction flask B . 6.25-cm. conical funnel tion process w a s C. Reagent grade granulated lead eliminated, because D . C.1a.a wool
uranium (VI) was reduced quantitatively to uranium (IV), l%hich could be titrated with a standard dichromate solution. This reductor has been found to be satisfactory for the indirect determination of sodium. Results given here show the percentage error obtained for weights of 3 to 6 mg. of sodium using a large reductor and 1 to 3 mg. of sodium with a small reductor. PROCEDURE
Large Reductor. -4liquot portions of a sodium chloride solution containing 3 t o 6 mg. of sodium were transferred to a beaker and sodium was precipitated as sodium zinc uranyl acetate by the method described by Iiolthoff and Sandell. The filter crucible containing the precipitate was placed in a Gooch funnel attached to a 250-ml. suction flask. The precipitate was dissolved in 50 ml. of an acid solution approximately 2.0 S with respect to sulfuric acid and 3.0 S with respect to hydrochloric acid (3.5 ml. of 36 LVsulfuric acid, 14.5 ml. of 12 S hydrochloric acid, and 40 ml. of distilled water). The crucible was washed thoroughly with 25-ml. portions of the acid mixture. The solution was passed through a lead reductor and received into a 5 % ferric sulfate solution, and the ferrous ions were titrated with a 0.05 potassium dichromate solution according to the method described (1). Results are given m Table I for 3 and 6 mg. of sodium.
Table I.
R e s u l t s Using L a r g e R e d u c t o r Error, %
Sodium Found, bIg. Wt. of sodium taken 6.02
=
6.00 mg.
6.01 6.01
Wt, of sodium taken = 3.06 mg. 3.07 3.06 3.07
+0.33 + O . 17 f0.17
+ O , 32 0.00 f0.32