The Analytical Approach Edited by Claude A.
Lucchesi
Figure 1. Analytical approach
Recognition and Solution of Production Problem in Vitamin Manufacture E. MacMullan, R. Hagel, and R. Gomez Quality Control Department Hoffmann-La Roche Inc. Nutley, NJ07110
It is a truism that the first step in solving an analytical problem is recognizing that you have a problem. This was demonstrated recently in the Roche Control laboratories during the testing of a lot of thiamine mononitrate. Thiamine (vitamin Bi) is an essential factor in human nutrition found naturally in rice hulls, cereal grains, yeast, liver, eggs, milk, and green leaves. It was originally isolated from rice bran; however, the thiamine used for dietary sup plements is almost entirely derived from chemical synthe sis. Thiamine mononitrate (I) synthesized at Roche must pass a rigid battery of chemical, physical tests to assure its purity and equivalence to the natural vitamin before being released for sale. "H3C.
Γ
XX
CR:
N(V CH3
In the course of this testing, the analyst observed that one lot did not pass the solution clarity test. At a concen tration of 2% in water, the solution was hazy whereas it should have been clear. In all other respects, the sample met specifications. At this point, the analyst could have simply failed the lot and returned it to Production for reprocessing where an ad ditional recrystallization would have undoubtedly brought the lot within specifications. But he recognized this as an unusual result worth a deeper investigation; therefore, he requested that the Analytical Research section look into the situation. Defining the Problem The general approach taken to define the problem was to: • Determine the level of impurity in the sample • Isolate some of the impurity • Consult with the production chemist on likely impuri ties • Subject the impurity to spectral and chromatographic analysis to determine its structure.
These steps are shown schematically in Figure 1. The purity of the lot was measured by phase solubility analysis with procedures similar to those described by Webb (1) and MacMullan (2). Phase solubility analysis is a technique for determining the purity of materials based on a careful measurement of their solubility behavior. The method has its theoretical origin in the Gibbs Phase Rule in which absolute purity is defined as single component be havior. A solid material in equilibrium with a saturated so lution which does not vary in composition as a function of the amount of excess solid in equilibrium with the solution is a pure solid. A solid whose saturated solution composi tion increases with increasing amounts of solid in equilibri um with the solution is impure to the extent of the increase in solution concentration. The phase solubility analysis indicated an impurity level of 1.3% (Figure 2). This changed the picture significantly. What had been considered a trace amount of insoluble im purity was in fact a significant amount of a slightly soluble impurity. This was a serious problem and made the identi fication of the impurity even more urgent. Identification of Impurity To obtain more of the impurity, 11 grams of the thi amine mononitrate was shaken with 550 ml of water for 30 min. The resulting suspension was filtered, and 49 mg of dried impurity was recovered for study by UV, IR, NMR, and mass spectrometry. At the same time, discussions with production chemists led to the information that the impu rity might be the immediate precursor in the synthesis (3), thiothiamine [3-(4-Amino-2-methyl-5-pyrimidinyl methyl)5-(2-hydroxyethyl)-4-methyl-4-thiazoline-2-thione] (II), which is only slightly soluble in water. H3CV J ^ H 2 N^ J 3 ^ ^CH2CH2OH
This apparently was confirmed when thin-layer chroma tography of the suspect lot showed an impurity spot at the ANALYTICAL CHEMISTRY, VOL. 47, NO. 4, APRIL 1975 · 473 A
„»o—o-
o_o.
»o—-°
I
f
!
•_. -
J
.- .u. ystem Composition, mg Sample/g Solvent
P;^llpM&£€!i#^^-ft^I:£:!>^
F i g u r e 2 . P h a s e solubility a n a l y s i s of t h i a m i n e m o n o n i t r a t e Sample: thiamine mononitrate lot A System: methanol, 20 hr @ 25°C Slope (computed by least squares with 9 5 % confidence): 1.30 ± 0.08% Extrapolated solubility: 2.63 ± 0.04 mg/g
F i g u r e 3 . P h a s e solubility a n a l y s i s of purified t h i a m i n e m o n o nitrate Sample: thiamine mononitrate lot A (phase purified) System: methanol, 20 hr @ 25°C Slope (computed by least squares with 9 5 % confidence): —0.08 ± 0.3% Extrapolated solubility: 2.92 ± 0 . 1 mg/g
same Rf as thiothiamine. However, when t h e plate was sprayed with thiochrome, t h e thiothiamine s t a n d a r d gave a fluorescent spot u n d e r longwave UV, b u t t h e impurity spot did not. F u r t h e r m o r e , t h e UV, IR, and N M R spectra were not consistent with t h e thiothiamine. T h e IR showed a carbonyl stretching b a n d between two strong electronegative groups, a n d t h e N M R showed chemical shifts and splitting p a t t e r n s consistent with a t h i a m i n e derivative containing a carbonyl in t h e 4-position of the thiazole ring. T h e identification d a t a are summarized in T a b l e I. A hypothetical structure of t h e impurity was obtained by replacing t h e sulfur in thiothiamine with a n oxygen. A search of t h e chemical literature showed t h a t this comp o u n d h a d been m a d e a n d was known by t h e trivial n a m e t h i a m i n e thiazolone [3-(4-Amino-2-methyl-5-pyrimidinyl methyl)-5-(2-hydroxyethyl)-4-methyl-4-thiazoline-2-one] (III). F u r t h e r m o r e , it had been prepared from t h i o t h i a m i n e by oxidation in alkaline m e d i u m (4).
Corrective Action
H,N,
H3a
CH,CH2OH
An a u t h e n t i c sample of thiamine thiazolone was prep a r e d by t h e literature procedure, a n d t h e structure was confirmed by IR, N M R , M S , and elemental analysis. In all cases, t h e spectra of t h e insoluble impurity were identical to t h e authentic thiamine thiazolone.
T a b l e I. S u m m a r y of S p e c t r a l a n d Chromatographic Results Technique
Results
Conclusions
Thin-layer chromatography UV s p e c t r u m
S i m i l a r Rf t o t h i o t h i a m i n e b u t different reaction to thiochrome Differs f r o m t h i o t h i a m i n e
Impurity not thiothiamine
IR s p e c t r u m
Carbonyl band between two strong electronegative groups Hypothesized Consistent with oxygen i m p u r i t y as analog of t h i o t h i a m i n e thiamine M o l e c u l a r i o n m/e 280 thiazolone Fragmentation consistent) w i t h oxygen a n a l o g of thiothiamine
NMR spectrum Mass spectrum
474 A ·
ANALYTICAL
CHEMISTRY,
VOL. 47, NO. 4, A P R I L
Once t h e impurity h a d been identified, t h e next step was t o institute procedures t o insure t h a t it would not appear in s u b s e q u e n t lots. F u r t h e r discussion with t h e production chemists indicated t h a t the p H h a d probably gone u p during t h e conversion of thiothiamine to t h i a m i n e mononit r a t e , allowing t h e formation of a small a m o u n t of thiamine thiazolone. T h e control steps instituted were: • B e t t e r p H control of t h e reaction • An in-process T L C done by t h e production chemist • P u r i t y determination of t h e finished lot by phase solubility analysis. Another necessary step was to determine t h e toxicity of t h e t h i a m i n e thiazolone. T h e LD50 in mice was greater t h a n 4000 mg/kg. T h i s is even less toxic t h a n t h i a m i n e which has a reported toxicity of 3000 mg/kg, and this represents a very low level of toxicity on an absolute basis (5). While we have instituted steps to eliminate this impurity, we have t h e a d d e d assurance of knowing it is a very nontoxic substance. In retrospect, the critical steps in t h e solution of this problem were: • Recognition of t h e problem by t h e control chemist • P u r i t y determination by phase analysis • Close cooperation between analytical a n d production chemists • Search of t h e chemical literature. Given t h e success of these four steps, t h e successful laboratory solution of t h e problem was almost inevitable. These steps in some form or other are fundamental to t h e solution of any production problem. P h a s e analysis h a d one more role to play in t h e process. T h i s m e t h o d can also be used as a separation technique since a t solution equilibrium all t h e impurities end u p in t h e solution phase, and t h e undissolved solids are essentially pure. T h i s " p h a s e purification" was r u n on a 50-gram sample of the i m p u r e lot a n d yielded 47 grams (94%) of material with a purity of 100.0% ± 0.1%. T h e analysis of t h e purified material is shown in Figure 3. T h i s procedure could be scaled u p for use in the production area.
References (1) T. J. Webb, Anal. Chem., 20,100 (1948). (2) E. A. MacMullan, paper presented at Land-O-Lakes Conference on Pharmaceutical Analysis, August 1969. (3) Maxion, U.S. Patent 2,844,579. (4) T. Matsukawa and H. Hirano, J. Pharm. Soc, Jap., 73, 379 (1953). (5) "The Merck Index," 8th éd., p 1037, 1968. 1975
Why is this the most popular syringe in the scientific world? ί|
• • • I ! 11 j ! • ' ι ' I " ι •• ' ! n ! 1111 ; 1111 ·. t ι · ι • \ 1111Ί11 > ι11 \ 11 ; η ^ " " " · · MK-am-mvt.
There are many reasons. Probably the foremost of which is its accuracy. We guarantee you absolute liquid delivery, accurate to within ± 1 % . . . and a repeatable accuracy to within ± 1 % . People who have used Hamilton syringes for almost two decades know that they can rely on that accuracy. That's a reputation we don't take lightly.
Each 700 Syringe is individually assembled and fitted by hand to exceedingly close tolerances. Attention to details The plunger is stainless steel that has been straightened by our precision machines and then centerless ground. We use alkali resistant borosilicate glass that is annealed after forming to relieve internal strains.
M rat
To create the most accurate bore possible, the glass is heatformed to a precision-ground mandrel. So, when the plunger is fitted to its barrel, you have a unique one-of-a-kind syringe, mated for precision. Consider all the options You have over 50 models and options from which to create your own special syringe: Seven different capacities, from 5 μΐ to 500 μΐ, needle lengths and gauges, luer tips, luer locks, plunger guides, and Chaney Adaptors.
One model is available with a handy removable needle, making it possible to replace bent or plugged needles. It locks securely in place, with virtually no dead volume. The price is right With inflation pushing prices up with every government bulletin,
CIRCLE 119 ON READER SERVICE CARD 476 A · ANALYTICAL CHEMISTRY, VOL. L. 47, NO. 4, APRIL 1975
"*T.r*9 ι * » · · » · · '
- > *
J
If
it's a pleasure to tell you of a product that has been the same price for 20 years. The basic 700 Syringe is and always has been priced at $18.00. As the costs, materials and labor have increased, we have been able to develop our syringe manufacturing skills and improve our speed of production to offset the higher costs.
Ui^— When you want quality and accuracy in a syringe . . . at 1950's prices . . . look for our trademark. It's become the symbol of the world's standard measuring device. Hamilton syringes are available from authorized dealers or direct from the factory. For more information and literature, write to John Nadolny, Hamilton Company, Post Office Box 10030, Reno, Nevada 89510.
ΗΑΜΙΙΧ©ΙΜ SYRINGES
