244 Aptalyst, April, 1968, Vol. 93,$j5. 24&248 Micro Determination of Carbonate in Dental Enamel BY J. A. WEATHERELL AND C. ROBINSON (Biological Research Unit, Dental School and Hospital, University of Leeds) A method is described for the micro determination of carbonate in 50-pg particles of human dental enamel. The technique is rapid and more sensitive than previous procedures. Amounts of carbonate from 0.5 to 3.0 pg have been determined with an accuracy of 4 to 7 per cent. (standard deviation). Carbon dioxide is liberated by dissolving enamel particles in acid. The gas forms a single bubble, flattened into a 100-p thick disc between the parallel glass surfaces of a Neubauer haemocytometer. The area of the flattened bubble is measured and the volume of gas calculated. By using this technique in combination with a recently developed sampling procedure, it has proved possible to measure variations in carbonate concentration within thin sections of dental enamel.RECENT investigations into the chemical composition of dental enamel necessitated the determination of micro amounts of carbonate. Previous studies of carbonate distribution1s2 have been carried out on pooled samples of powdered enamel ground from large numbers of teeth. It was found that human enamel contained between 2 and 3 per cent. of carbonate, expressed as carbon dioxide, and that its concentration increased from the surface to the interior of the enamel. Pooled samples cannot, however, provide information about variations within the single tooth, and the aim of the present investigation was to evolve a method by which this could be done.A recently developed techniqueS was used to dissect the enamel of sections of teeth just over loop thick, into about one hundred particles, each weighing 20 to 50pg. A dissected section is shown in Fig. 1. The carbonate content of each particle, expressed as carbon dioxide, would be in the order of 1 pg. No technique was available for determining such a small amount as chemical methods were too insensitive, and most physical procedures could not be adapted easily to the manipulation of the small amounts of carbon dioxide derived from microgram amounts of carbonate. A technique has now been developed, based on a method described by Krogh in 1908* and by Lewis and Lippold in 1956.6 It has proved possible to liberate carbon dioxide as a single, stable bubble from the carbonate present in enamel.The volume of the bubble can be measured and the carbonate concentration of the particle determined. This paper describes the procedure, and presents some results showing the distribution of carbon within a 100-p thick section of human enamel. EXPERIMENTAL FORMATION OF A SINGLE, STABLE CARBON DIOXIDE BUBBLE- Aqueous solutions of strong acids readily dissolve dental enamel, liberating carbon dioxide from the mineral. Initial attempts to measure the carbon dioxide evolved were hampered by two technical problems. Firstly, acid alone produced many separate bubbles (Fig. 2a), the total volume of which could not be measured, but this difficulty was surmounted by adding ethanol. With an aqueous solution containing 18 per cent.w/v of perchloric acid and 40 per cent. v/v of ethanol a single measurable bubble was produced (Fig. 2b). When, occasionally, two or three bubbles formed, slight manipulation was sufficient to make them coalesce. The second difficulty arose from the shrinking of the bubbles as the carbon dioxide dissolved in the acid. To avoid this, the acid - ethanol mixture was pre-saturated by bubbling carbon dioxide through it. Under the conditions described below the single bubble of gas remained stable for 1 hour or more. 0 SAC and the authors.Fig. 1. Dissection of enamel in 100-p thick tooth section (from Weatherell, Weidmann and Hanim7) ( a ) (L-) Fig. 2 . separate bubbles forming in aqueous acid ( x 60) ; single bubble produced in acid - ethanol mixture by about 1 pg of carbon Formation of gas bubble from enamel particle: (a) (b) d 1 oxide To face page 244WEATHERELL AND ROBINSON 245 A = Particle of sample under cover-slip B = Central platform of haemocvtometer.F = Perspex lid resting on chamber show- G = Lead blocks ing lOO$ wide gap betwekn platform and H = Photographic enlarger screen cove r-s I i p I = Enlarger lens C = Optically plane cover-slip J = Light source D = Brass chamber K = Water cell to absorb heat from light source El= Vessel containing acid - ethanol mixture L = Plane mirror Fig. 3. Diagram of apparatus and haemocytometer APPARATUS- The apparatus is shown in Fig. 3. The particle, A, is placed in the central cell of a Neubauer haemocytometer.This has been illustrated, both apart from the apparatus and also in the position it occupies during a determination. The haemocytometer is a glass slide normally used for blood-cell counts, with a central platform, B, set 100 1 p lower than the adjacent surfaces. An optically plane cover-slip, C, lies across the platform, forming a 100-p deep space. When the cover-slip is in close contact with the raised outer faces of the haemocytometer, interference fringes appear between the raised surfaces and the cover-slip. The slide is placed within a brass chamber, D, the base of which is maintained at 22" C by a flow of water. A stream of carbon dioxide can also be passed through this chamber. The acid-ethanol is saturated with carbon dioxide in a vessel, E, from which small amounts of the mixture may be quickly removed.The acid is transferred from this vessel to the haemocytometer by inserting a capillary tube through the hole in the Perspex lid, F, of the brass chamber, and is pulled into the space between slide and cover-slip by capillary action to fill the 100-p. wide cell containing the particle. The front of the acid would sweep out the enamel specimen, but as the particles are dissected from sections slightly more than 100-p thick, they are held in place by the cover-slip. Because of this, the cover-slip is slightly bowed over the particle of enamel, being weighted at its edges by small lead blocks, G. As the enamel dissolves, the cover-slip flattens out. The entire process of enamel dissolution and bubble formation is observed with the ground-glass screen of a photographic enlarger, H, on to which an image of the particle and bubble is projected with a standard magnification.The area of the bubble is measured and the volume of gas determined, as described below. The determination is extremely rapid, as the time required for the bubble to form is rarely more than 5 minutes.246 WEATHERELL AND ROBINSON : MICRO DETERMINATION [AutdySt, VOl. 93 PROCEDURE- Before carrying out a series of determinations, saturated the acid - ethanol mixture with carbon dioxide by bubbling the gas through it at a rate of about 500ml per minute for about 1 hour. Place the specimen in the centre of the Neubauer haemocytometer, position the cover- slip over it and across the 100-,u deep cell.Lay the slide on the thermostatically controlled base of the brass chamber and place the two lead blocks carefully on the edges of the cover- slip. Put the Perspex lid on the brass chamber and position the specimen beneath the photographic enlarger. By turning the 3-way tap, flush the chamber with a stream of carbon dioxide for about 90 seconds. The gas flow, at a rate of about 500 ml per minute, must be maintained until the determination is complete. Remove about 2 0 4 of acid- ethanol by glass capillary or micropipette from the saturation vessel and introduce it under the Neubauer cover-slip. Attention should meanwhile be concentrated on the magnified image of the specimen, which is focused on the ground-glass screen of the photographic enlarger, watching carefully to ensure that none of the gas evolved from the specimen is swept by the acid from below the cover-slip.Observe the formation of the bubble, or bubbles of gas until the enamel particle is completely dissolved. The time of dissolution varies from about 40 seconds to 5 minutes, depending on the size of the particle and the solubility of the enamel. If more than one bubble has formed, a slight pressure on the Perspex lid, which just rests on the two lead blocks, is usually sufficient to cause the bubbles to coalesce. To ensure satisfactory bubble formation it is essential'to maintain a high standard of cleanliness. The haemocytometer and cover-slip should be kept in chromic acid between determinations. Provided this is done, interference fringes will have formed between the cover-slip and the raised portions of the slide when the determination is complete.Under these conditions the depth of the Neubauer well is 100 The area contained within the outer edges of the meniscus of the bubble is measured by tracing round its magnified image and determining the area of the tracing. The total volume of the bubble is determined as described below. ASSESSMENT OF BUBBLE VOLUME- In calculating the volume of the bubble from the area enclosed by the outer edge of the meniscus, a correction must be made for the error caused by the curve of the meniscus itself. Lewis and Lippolds made an approximate correction by assuming the meniscus to be semi-circular in cross-section. In the present experiments the apparatus was calibrated directly.The shape of the bubble was assumed to be similar to that of a flattened drop of mercury. Small weighed drops of mercury, the volume of which could be calculated from their known weight, were placed beneath the cover-slip of the haemocytometer. The haemo- cytometer was placed in the brass chamber and the areas of the flattened drops measured by drawing round their projected images. This correlation between area and volume gave results similar to those derived from calculation, but provided at the same time a convenient means of making direct comparisons between one haemocytometer and another, and of checking the effect of any alteration in the geometry of the apparatus on the estimated volume of the bubble. CORRECTIONS FOR TEMPERATURE AND PRESSURE- The brass chamber in which the haemocytometer slide was placed was maintained at 22' C by water circulating through its hollow base.The bubble was assumed to be formed at atmospheric pressure. Any small effect the surface tension of the narrow gas - liquid interfaces might have on the internal pressure of the bubble was ignored. This was justified by the calibration results shown in Table I. The weight of carbon dioxide was calculated after correction of the carbon dioxide volume to standard conditions of temperature and pressure. CALIBRATION - The material had to be soluble in the acid - ethanol mixture, possess a carbonate content similar to that of enamel, be chemically homogeneous and sufficiently stable, especially with regard to carbon dioxide and water, to permit its use as a weighable standard substance.1 p. CALCULATION It proved extremely difficult to find a suitable calibration substance.April, 19681 OF CARBONATE IN DENTAL ENAMEL 247 These criteria were not fulfilled by any of the numerous inorganic or mineral carbonates considered, or by many artificially prepared fused materials. The most suitable calibration substance was an amorphous calcium - phosphate - carbonate precipitate prepared according to the precipitation diagram of Bacchra, Trautz and Simon.6 The material was compressed into a thin disc and its carbonate content determined both by micro diffusion with Conway cells and also by using the bubble technique. For the less sensitive micro-diffusion procedure it was necessary to use about fifty times the weight of sample used in the bubble technique.The small particles of compressed precipitate tended to effervesce vigorously. Thus to prevent a loss of carbon dioxide, the particles wqe moistened with a tiny spot of glycerol, which appreciably slowed their rate of dissolution. The results are shown in Table I; the accuracy of the present method appears to be between 4 and 7 per cent. (standard deviation). AS some of this error undoubtedly arose from slight differences between the particles of compressed precipitate, the determination itself may well be more accurate than this. TABLE I COMPARISON OF CARBON DIOXIDE CONTENT OF THE CALIBRATION MATERIAL, AS DETERMINED BY THE CONWAY MICRO-DIFFUSION METHOD AND BY THE BUBBLE TECHNIQUE Conway micro-diffusion method I A \ Precipitate Number of used determinations Mean result A 6 4.49 f 0.10 Standard deviation = 2.2 per cent.B 6 3.16 f 0.11 Standard deviation = 3.5 per cent. C 6 2-69 5 0.09 Standard deviation = 3.3 per cent. Bubble technique r \ A Number of determinations Mean result 6 4.48 f 0.17 Standard 5 2.86 f 0.12 Standard 12 2-76 f 0.19 Standard deviation = 3.8 per cent. deviation = 4.2 per cent. deviation = 6.9 per cent. RESULTS Some results obtained by the analysis of single enamel particles, each weighing between 20 and 50 pg, are shown in Fig. 4. Three of the curves shown (open circles) were obtained by determining the carbonate content of three columns of contiguous particles dissected from a single section. The carbonate distribution is essentially similar to the curve obtained by Little and Brudevoldl (closed circles) by determining the carbonate in pooled samples of enamel ground in layers from large numbers of teeth.DISCUSSION The technique described is considerably more sensitive than previous methods of car- bonate determination, enabling amounts of carbon dioxide of the order of 1 pg to be deter- mined. The procedure is straightforward, the apparatus simple, and analyses can be carried out rapidly. It has made possible a study of carbonate distribution in thin sections of dental enamel. The results presented in Fig. 4 show that the carbonate concentration was lower in surface regions of the tooth section than in the enamel interior. This distribution is essentially similar to that described by Little and Brudevoldl- and by Nikiforuk and Graingers; from the analysis of pooled enamel.The absolute values obtained by these workers for the carbonate content of the surface regions agree well with those of the present study, although their results for samples taken from the tissue interior were lower. The difference may be partly attributable to biological variation, or could be ascribed to the ability of the sampling technique used in the present work to assess more accurately the area of sampling without danger of dentinal contamination. The method was specifically designed for studying dental enamel and cannot be applied directly to the analysis of other mineralised tissues. In enamel, the small amount of organic material present, usually less than 0.5 per cent., dissolves or disintegrates and so does not interfere with the determination.The 20 per cent. of collagenous matrix present in bone or dentine, however, hinders dissolution of the mineral and mechanically prevents the248 WEATHERELL AND ROBINSON 10 20 30 40 50 60 70 80 90 I Depth from surface to dentino- enamel junction, per cent. Fig. 4. Distribution of carbonate from surface to dentind-enamel junction in human dental enamel: 0, results from three series of particles from one 100-p thick section of enamel; and 0, results obtained by analysis of pooled enamel powder (from Little and Brudevold’) formation of a measurable bubble. Preliminary tests suggest that this difficulty can be overcome by removing the organic material before the determination of carbonate. Small particles of bone or dentine were refluxed overnight in a 3 per cent.solution of potassium hydroxide in ethylene glycol. The specimens were thoroughly washed with methanol to remove all traces of the alkaline reagent. Carbonate was then determined in the de-proteinised material that remained. The results obtained were similar to previous estimates of the carbonate concentration in bone and dentine. If the gas bubble is absorbed in 0.1 N sodium hydroxide, a small amount of gas, constituting about 1 per cent. of the original bubble volume, invariably remains, the nature and origin of which is unknown. There is also uncertainty about the true internal pressure of the bubble. In the light of the calibration results, however, these factors do not appear to be significant. The authors are indebted to the Medical Research Council for a grant in support of this work, and to Mr. G. Naylor for technical assistance. Certain questions remain. REFERENCES 1. 2. Little, Marguerite F., and Brudevold, F., J . Dent. Res., 1958,37, 991. Nikiforuk, G., and Grainger, R. M., in Stack, M. V., and Fearnhead, R. W., Editors, “Tooth Enamel,” A report of the Proceedings of an International Symposium on the Composition and Fundamental Structure of Tooth Enamel, London Hospital Medical College, 1964, John Wright and Sons Ltd., Bristol, 1965, p. 26. Weatherell, J. A., Weidmann, S. M., and Hamm, Stella M., A r c h Oral Biol., 1966, 11, 107. Krogh, A., Scund. Arch. PhysioE., 1908, 20, 279. Lewis, H. E., and Lippold, 0. C . J., J . Scient. Instrum., 1956, 33, 254. Bachra, B. N., Trautz, 0. R., and Simon, S. L., Adv. Fluorine Res. Dental Caries Prevention, Weatherell, J. A., Weidmann, S. M., and H a m , Stella M., Caries Res., 1967, 1, 42. 3. 4. 5. 6. 7. 1965, 3, 101. Received September 4th, 1967