- Email: [email protected]
A sorbent tube for oral malodour monitoring
Talanta 62 (2004) 421–426
A sorbent tube for oral malodour monitoring Julio Rodr´ıguez-Fernández a , Regina López-Fernández a , Rosario Pereiro a , Manuel Menéndez b , José Mar´ıa Tejerina b , Alberto Sicilia b , Alfredo Sanz-Medel a,∗ a
Department of Physical and Analytical Chemistry, University of Oviedo, Julián Claver´ıa 8, 33006 Oviedo, Spain b Department of Estomatology, University of Oviedo, Catedrático José Serrano, s/n 33006 Oviedo, Spain Received 9 May 2003; received in revised form 25 July 2003; accepted 13 August 2003
Abstract Volatile sulphur compounds (VSCs) and particularly hydrogen sulphide are considered as the predominant gases causing oral malodour. In this paper, a simple alarm sensor has been developed for VSCs determination in mouth air. The device consists of a glass tube packed with a solid sensing phase. The VSCs react with the sensing phase to produce a change in the colour of the sensor visible with a naked eye. Different “reagents” were investigated to develop the sensing phase (neocuproine + Cu(II), bathocuproine + Cu(II), resazurin, 2,6-dichlorophenolindophenol and lead acetate), finding the neocuproine + Cu(II) as the best for our purposes. Also, different substrates such as Amberlite XAD-4 and XAD-7 and different trademarks of silica gel were tested as solid supports, being selected the silica gel. A device consisting of a glass tube packed with the sensing phase was optimized and tested with halitosis patients as a rapid illness test and the results compared with those obtained with a commercially available instrument, the HalimeterTM , used for the determination of VSCs in mouth air. The results exhibited acceptable agreement between the proposed “qualitative” alarm sensor and a commercially available technique selected as reference, showing the possibility of using this “visual sensor” to control the halitosis and its evolution with an eventual treatment, by the own patient. © 2003 Elsevier B.V. All rights reserved. Keywords: Visual sensor; Volatile sulphur compounds; Halitosis; Neocuproine–Cu(I)
1. Introduction Halitosis is a general term used to describe unpleasant odour coming out from the mouth air and breath, independently from the source where the odour substances are coming from [1–5]. When the sources are oral it is termed “oral halitosis”, being the most frequent source of breath malodour in healthy subjects. Oral pathogenic halitosis is a common problem affecting the adult population [6,7]. It has been previously reported that volatile sulphur compounds (VSCs), such as methyl mercaptan and, in particular, hydrogen sulphide are the major gases associated with oral malodour [8,9]. These odours arise from microbial degradation from proteins (especially these that contain cysteine and methionine), peptides and aminoacids that are present in saliva, gingival crevicular fluid or in food that is retained on the teeth [10]. ∗ Corresponding author. Tel.: +34-98-510-3474; fax: +34-98-510-3125. E-mail address: [email protected] (A. Sanz-Medel).
0039-9140/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2003.08.014
In some studies [11,12], the relationship between the amount of volatile sulphur compounds and the severity of periodontal disease, or oral prophylactic conditions related to the accumulation of tongue coating, gingival fluid, and saliva putrefaction, has indicated that septic and unsanitary oral conditions are often the major cause of apparent halitosis. For the assessment of oral malodour, two types of methods are generally used: one is the subjective organoleptic examination [13] and the other is the quantitative measurement of VSCs; e.g. using gas chromatography [6,14]. Subjective organoleptic examination by nasal sniffing is the simplest and commonly used approach to directly evaluate expelled mouth air. Although this method has been considered as a reference standard of oral malodour measurement, it raises several problems such as the variation between judges on the ranking of the same sample [15]. On the other hand, quantitative measurement by gas chromatography offers consistent results, but such measurements show several practical problems to be used by the patient or the clinician itself (e.g. too complex system, technical expertise
422
J. Rodr´ıguez-Fern´andez et al. / Talanta 62 (2004) 421–426
required, and procedures are costly and time-consuming [6,11]). Due to the growing interest of this topic in the medical field, new techniques for measurement of halitosis related sulphides have been reported in recent years [16–18]. In most cases, these techniques are based on redox measurements and include a zinc-dioxide semiconductor sensor [16], a tin-oxide semiconductor gas sensor [17] (the BB ChekerTM , Tokkuyala Corp., Tokyo, Japan), and the HalimeterTM developed by Interscan Corporation (Chatsworth, CA 91313-2496, US). The last one seems to be the most common instrument for monitoring sulphide-type compounds in the oral cavity. More recently, it has been commercialised an amperometric sulphide silver sensor (Diamond General Development Corp., Michigan, US) which permits the measurement of VSCs in localised regions inside the mouth. To our knowledge only two optical sensors has yet been developed for measurement VSCs in mouth air [19,20]. One is based on the measurement of the fluorescence produced by the reaction of the complex Hg(II)–2,2 pyridylbenzimidazole (immobilised on a silica gel support) with the sulphide ion. The other is a reversible optical fibre sensor based on reflectance measurements of the 2,6-dichlorophenolindophenol on silica gel. All the techniques above mentioned for VSCs monitoring are addressed to the use by the dentist itself, because they require instrumentation more or less sophisticated. Thus, the objective of this work was the development of a simple, cheap and easy to handle “visual sensor” to be used by the patients for the autocontrol of their halitosis in an easy way and without the need of any instrument. For this purpose, different chemicals such as neocuproine + Cu(II), bathocuproine + Cu(II), resazurin, 2,6-dichlorophenolindophenol and lead acetate have been assayed as reagents. Also, different solid supports (Amberlite XAD-4, XAD-7 and silica gel) have been investigated to immobilise the reagents to obtain an adequate “active” phase, changing its colour when VSCs concentration in mouth air surpassed a threshold value. After optimisation for best perception of the colour change, an “active” phase has been selected and evaluated for the detection of halitosis in 72 subjects. The results were compared with those obtained with an electrochemical instrument commercially available (HalimeterTM ) for halitosis monitoring.
drate were purchased from Merck (Darmstadt, Germany). Methanol used to dissolve reagents, was purchased from Prolabo S.A. (Barcelona, Spain). Gallocyanine, toluidine blue O and Congo red, employed as colorants, were purchased from Aldrich. Sodium acetate, sodium tetraborate, sodium hydrogencarbonate, tris(hydroxymethyl)aminomethane, acetic acid, hydrochloric acid, boric acid and sodium hydroxide, all of them from Aldrich, were employed to prepare buffers in the interval from pH 3 to 12. To prepare the neocuproine + Cu(II) and the bathocuproine + Cu(II), the appropriate volumes of organic reagents and copper dissolved in methanol were mixed keeping a molar rate neocuproine: Cu(II), bathocuproine: Cu(II) of 3:1. Amberlite XAD-4, Amberlite XAD-7, silica gel Davisil grade 646 and silica gel Merck grade 10181 purchased from Aldrich and silica gel 40 purchased from Merck were investigated as solid supports. TedlarTM (polyvinyl fluoride) sampling bags with polypropylene valves (Supelco, Madrid, Spain) were used for analyte gas dilution and also for testing the different sensing phases prior to use with patients. Hydrogen sulphide and methyl mercaptan gases were selected as “model” VSCs. They were purchased from Carburos Metalicos (Oviedo, Spain) and were of analytical reagent grade. 2.2. Instrumentation Since visual detection of the analyte was aimed to, no optical instrumentation was used for the sensor. For the preparation of the sensing phase a rotaevaporator Helidoph VV2000 with a water bath Helidoph WB2000 and a pHmeter with a glass electrode (model MicropH 2000, Crison Instruments S.A., Alella, Spain). PyrexTM minicolumns of 8 mm o.d., 3 mm i.d. and 20 mm length were filled with the solid sensing phase (Fig. 1). A common plastic mouthpiece (as those used for tobacco smoking) was inserted at an end of the minicolumn. A 250 ml plastic bag was inserted at the other end. For each test, the patients should blow from the mouthpiece through the minicolumn up to the 250 ml plastic bag is filled,
2. Experimental 2.1. Reagents and solutions All the colorimetric reagents were of analytical grade. Neocuproine hydrochloride hydrate, bathocuproinedisulfonic acid disodium salt hydrate and resazurin were purchased from Sigma–Aldrich (Madrid, Spain), 2,6-dichlorophenolindophenol, lead acetate and copper sulphate pentahy-
N
N
CH3
CH3
2
Cu (I) Fig. 1. Chemical structure of the neocuproine–Cu(I).
J. Rodr´ıguez-Fern´andez et al. / Talanta 62 (2004) 421–426
thus meaning that such air volume has passed through the minicolumn. A commercially available instrument for screening halitosis (HalimeterTM ) was selected for comparison. 2.3. Preparation of the solid sensing phase Methanolic solutions, containing the investigated reagent and the required chemicals as well as the solid support, were evaporated in a rotaevaporator with a water bath at 20 ◦ C (solutions were prepared in methanol). After the rotaevaporation, the solid phase is washed with a convenient solution (when required) and dried at room temperature. Finally, the sensing phase is stored in the refrigerator at 4 ◦ C and in the darkness.
3. Results and discussion 3.1. Selection of the colorimetric reagent For the development of the solid sensing phase different reagents were evaluated: neocuproine + Cu(II) [21,22], bathocuproine + Cu(II) [23], lead acetate [24], resazurin [25], and 2,6-dichlorophenolindophenol [26]. All of them were selected from the literature due to their good analytical performance characteristics for hydrogen sulphide analysis. The solid “active” phases are packed into the glass minicolumn to produce the visual sensor. Then, a known quantity of H2 S gas is introduced into the minicolumn until a change is observed in the colour of the sensing phase and this point is defined as the “visual detection limit” (visual D.L.). Results obtained for the silica gel 40 are shown in Table 1, where it can be seen that the best “visual D.L.” are obtained for neocuproine–Cu and bathocuproine–Cu reagents. Since the analytical behaviour of both chelates proved to be quite similar, being bathocuproine price higher, neocuproine was eventually selected for the further development of an adequate VSCs sensing phase. 3.2. Reagent immobilisation Five different solid supports were compared for the immobilisation of the neocuproine + Cu(II): two nonionic polyTable 1 Selection of the coloured reagent (solid support: silica gel 40) Reagent
Experimental conditions (silica gel concentration)
Visual D.L. (ppb (v/v)) (H2 S)
Neocuproine–Cu(II) Bathocuproine–Cu(II) Resazurin
0.44 g/g 0.055 g/g 0.4 mg/g; Cu(II): 0.16 mg Cu/g 5 mg/g 0.02 g/g
250 250 1000
Lead acetate 2,6-Dichlorophenolindophenol
2500 500
423
Table 2 Selection of the solid support (using the complex neocuproine–Cu(II)) Solid support
Visual D.L. (ppb (v/v)) (H2 S)
Amberlite Amberlite Silica gel, Silica gel, Silica gel,
500 1000 250 250 250
XAD-4 XAD-7 DavisilTM , grade 646 Merck, grade 10181 grade 40
meric adsorbents (Amberlite XAD-4 and Amberlite XAD-7) and three different silica gel supports (Davisil grade 646, Merck grade 10181 and silica gel grade 40). Solid sensing phases were prepared as previously described. Five identical neocuproine + Cu(II) methanolic solutions were prepared by dissolving 0.055 g of the reagent in 50 ml of methanol and adding 10 ml of 7.45 × 10−3 M CuSO4 . Then, to each of the solutions was added 1 g of the corresponding solid support and carried out to dryness into the rotaevaporator at 20 ◦ C. Finally, the solid phases were washed with distilled water up to clear washing liquids and dried at room temperature. Sensing phases prepared in this way were packed into the glass minicolumn and their performance was assessed with hydrogen sulphide samples prepared by dilution of the appropriate H2 S (g) in a TedlarTM bag containing wet air. All silica gel supports (silica gel 40, Davisil grade 646 and Merck grade 10181), offered good visual detection limits according to our purposes (Table 2), but the stability of the reagent in the silica gel Merck grade 10181 and the Davisil grade 646 was worse than in silica gel 40. Therefore, silica gel 40 was selected as the solid support. Similar detection limits were obtained for hydrogen sulphide than for methyl mercaptan. 3.3. Neocuproine + Cu(II) (effect of reagent concentration) The reaction of the neocuproine + Cu(II) with hydrogen sulphide is a redox process where a change in the colour is produced due to the reduction of the complex by the sulphide ion [21,22]. Neocuproine has been traditionally used with the aid of reducing agents for the determination of Cu [27]. Neocuproine + Cu(II) reacts with sulphide ions (Fig. 1) producing a complex with a different colour as a result of the reduction of the Cu(II) to Cu(I) and this reaction can be used for the determination of H2 S [21]. To avoid complex dissociation from the copper bischelate to the monochelate, the molar ratio neocuproine:Cu has to be 3:1. Different concentrations of neocuproine + Cu(II) immobilised on silica gel 40 were tested. It was observed that for concentrations in the interval 0.44–0.055 g of neocuproine (and the Cu(II) in the proper amount for obtaining a 3:1 ratio) per gram of solid support, there was no significant changes in the “visual detection limit” observed. On the other hand, the stability of the sensing phases increased at lower concentrations. Thus, 0.055 g of neocuproine per
424
J. Rodr´ıguez-Fern´andez et al. / Talanta 62 (2004) 421–426
gram of the solid support was eventually selected providing a net colour change for 250 ppb (v/v) of H2 S (g) concentration in air. The stability of this “active” phase was at least 3 months storage (in the refrigerator at 4 ◦ C and so in the darkness). 3.4. pH effect in the preparation of the sensing phase Using 0.055 g neocuproine + 10 ml 7.45×10−3 M CuSO4 per gram of silica gel 40, the optimisation of the pH of the sensing phase was investigated by washing it with different buffers and then, drying it again at room temperature. Different pHs were assayed (3, 4, 5.4, 7.5, 8.5 and 12), but neither the sensitivity nor the stability of the sensing phase were observed to be affected. Therefore, the sensing phase was just washed with distilled water in subsequent experiments. 3.5. Effect of a “screen” dye addition When neocuproine + Cu(II) reacts with hydrogen sulphide in the solid support, the chelate is transformed into neocuproine−Cu(I), changing its colour from slight green to yellow. Since we are using “visual detection”, the colour changes are not always easily perceptible for low concentrations of analyte. For this reason, different “screen” dyes were assayed (in order to change the perceived colours before and after the reaction, making the visual change easier to be detected). These “screen” dyes work like optical filters making the colour changes, produced by the presence of the hydrogen sulphide, more perceptible. The assayed dyes were: resazurin (violet), toluidine blue O (blue), gallocyanine (blue) and Congo red (red). These dyes were added to the mixture before rotaevaporation, and then carried out to dryness following the procedure described in Section 2.3. No enhancement of the perception of the colour changes with single or different mixtures of the above dyes was observed.
3.6. Final design of the sensor The final design of the sensor sought should give rise to a robust and simple device as such visual sensors are going to be used by people who have not received special training for their use. Also, the device should be cheap to help the commercialisation of the product. Attending these reasons, two approaches were assayed: glass minicolumns filled with the sensing phase [28] and paper sticks with the sensing phase on the surface of the paper [29]. After several experiments, the approach of using minicolumns filled with the sensing phase was selected. Different tubes (1.25, 2.0, 3.0, 4.0 mm i.d.) were assayed looking for a good “visual detection limit”. The used tubes were made of PyrexTM with thick walls (8 mm o.d.) to obtain a visual effect similar to a lens. The best analytical results were obtained with the tube with thinner internal diameter. Unfortunately, on using such devices the patients had problems blowing through this thin tube (due to the pressure created inside the tube by the sensing packed material), therefore, a tube with 3.0 mm i.d. was eventually chosen as a compromise between good VSCs detectability and easy blowing through it. The sensing phase proved to be non reversible over a period of 24 h after its use. To optimise the perception of the colour change, the sensing phase was eventually placed between two slices of white silica gel (trying to get a good contrast between the colours) within the minicolumn (see Fig. 2). 3.7. Interference studies The interference studies were carried out by simulating, as much as possible, the behaviour of possible interferent compounds present in the mouth air and breath. Different samples of mouth air were sampled from several volunteers without halitosis problems (control) under different conditions and evaluated with the “visual sensor” developed. No colour changes were observed. The samples were doped with 300 ppb (v/v) of H2 S (g) and different conditions of mouth
Filters
2.5 mm
3 mm
250 mL bag 250 mL bag
White silica gel
Mouthpiece
Sensing phase
Fig. 2. Schematic representation of the sensing system, including the mouthpiece and the plastic bag.
J. Rodr´ıguez-Fern´andez et al. / Talanta 62 (2004) 421–426
air were studied: after smoking a cigarette, after cleaning the tooth with a fluoride tooth paste, after cleaning the tooth with an elixir, after eating chewing gum with chlorophyll and, finally, after drinking alcohol. No difference between the samples with and without the potential interference was noticed. Therefore, such common mouth air environments can be considered not disturb the “visual sensor” performance. 3.8. Analysis of samples from halitosis patients For the validation of the proposed “alarm test”, comparative tests were carried out with halitosis patients and with volunteers without halitosis problems at the Department of Estomatology at the University of Oviedo. In all cases, the “visual sensor” results obtained by the dentists were compared with those obtained in parallel with an alternative quantitative technique (HalimeterTM ) used for malodour monitoring.
425
Comparative studies carried out in close collaboration with the Department of Estomatology from the University of Oviedo/University of Granada, indicated that a “socially accepted” VSCs limit in mouth air (organoleptic tests versus Halimeter) could be taken as about 250 ppb (v/v) [30]. As the “visual” sensor developed is a rather qualitative device (“alarm sensor”) there are only two possible lectures: positive (+) when the level of the VSCs in mouth air is superior to 250 ppb (v/v) (halitosis patient), and negative (−) when the value is lower than 250 ppb (v/v). Table 3 shows a comparison between the two methods of the values obtained for the control volunteers without halitosis problems and patients. As can be seen, a good correlation between the results obtained with our “visual” sensor and those obtained with the HalimeterTM was observed. From the results, a 2.8% of false negatives and a 5.6% of false positives were obtained. The percentage reliability according to Valcarcel [31] (reliability(%) = 100 − false negatives (%) − false positives (%)) was 91.6% (commonly, reliability
Table 3 Mouth air detection Volunteers
HalimeterTM values
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
119 215 119 262 102 190 118 127 156 120 132 128 177 172 96 163 158 169 1034 453 425 757 348 280 232 1658 542 408 1673 220 258 187 677 380 309 341 ∗
Unexpected results.
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
10 17 10 21 8 15 9 10 12 10 11 10 14 14 8 13 13 14 62 27 26 45 21 17 20 99 33 24 100 13 15 11 41 23 19 20
Visual sensor values
Volunteers
HalimeterTM values
− − − − − − − − − − − − − − − − − − + + + + + + + + + + + +∗ + +∗ + + + +
37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
96 261 250 191 121 234 210 110 223 234 220 295 193 208 329 209 154 258 300 928 136 115 491 240 1323 435 1401 886 752 342 252 546 256 260 546 309
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
8 21 20 15 10 19 17 9 18 19 18 24 15 17 26 17 12 21 18 56 8 7 29 14 79 26 84 53 45 21 15 33 15 16 33 19
Visual sensor values − − − − − − − − − − − −∗ − − −∗ − − − + + +∗ +∗ + + + + + + + + + + + + + +
426
J. Rodr´ıguez-Fern´andez et al. / Talanta 62 (2004) 421–426
percentages lower than 80% are considered as unacceptable [31]). Therefore, the results of Table 3 demonstrate the utility of the proposed “visual sensor” as an “alarm test” for the direct control of the halitosis by the own patient. 4. Conclusions An “alarm” visual sensor for oral malodour is proposed. The developed “visual sensor” is based on the colour change of a sensing phase of neocuproine–Cu and this device offers the possibility of daily control of halitosis problems by the patient himself. The proposed device is very simple to use and does not require any sample pre-treatment of mouth air. The non reversibility of the chemical system should not constitute a problem, because the sorbent tube was thought as something cheap of just one use (for example, to include a set of them into any already available commercial kit for bad breath cleaning). Besides, the production of the sensing phase is very easy. In any case, it should be taken into account that although neocuproine could be considered as a hazardous compound, this should not be a problem because it does not get in contact with the potential user. In any case, a quality control must be carried out in an eventual massive routine preparation of the sorbent tubes. Acknowledgements The authors thank Dentaid S.A. (Cerdanyola, Spain) for lending the HalimeterTM for mouth air determinations. Financial support from FICYT (Spain) and Dentaid S.A. (Cerdanyola, Spain) through the cooperative project number PA-SAL-97-01 is gratefully acknowledged. Thanks are particularly given to Mr. Enric Masdevall and Mr. Jaime Arum´ı for their continuous help and interest. Also, financial support from “Plan Nacional de I + D” (Spanish Ministry of Science and Technology) through the project Ref. PPQ2000-1291-C02-02 is gratefully acknowledged. References [1] M. Sanz-Alonso, ROE 1 (1996) 97. [2] M. Rosenberg, C. McCulloch, J. Periodontol. 63 (1992) 776.
[3] K. Yaegaki, J.M. Coil, J. Can. Dent. Assoc. 66 (2000) 257. [4] M. Quirynen, J. Clin. Periodontol. 30 (5) (2003) 17. [5] J. Rodr´ıguez, R. Pereiro, A. Sanz-Medel, Spectrosc. Eur. 14/2 (2002) 6. [6] J. Tonzetich, J. Periodontol. 48 (1977) 13. [7] M. Shimura, S. Watanabe, M. Iwakura, Y. Oshikiri, M. Kusumoto, K. Ikawa, S. Sakamoto, J. Periodontol. 68 (1997) 1182. [8] J. Tonzetich, Archs. Oral Biol. 16 (1971) 587. [9] N.F. Schmidt, S.R. Missan, W.J. Tarbet, A.D. Cooper, Oral Surg. Oral Med. Oral Pathol. 45 (1978) 561. [10] W.J. Loesche, E.H. DeBoer, in: D. Van Steenberghe, M. Rosenberg (Eds.), Bad Breath: A Multidisciplinary Approach, Leven University Press, 1996 (Chapter 3). [11] Y. Yasuno, M. Iwakura, Y. Shimada, J. Dent. Health 39 (1998) 663. [12] H. Miyazaki, C. Fujita, I. Soh, T. Takewara, in: D. Van Steenberghe, M. Rosenberg (Eds.), Bad Breath: A Multidisciplinary Approach, Leven University Press, 1996. [13] M. Rosenberg, G. Kulkany, A. Bosy, C. McCulloch, J. Dent. Res. 70 (1991) 1436. [14] M. Solis-Gaffar, H. Neiles, W.C. Rainieri, R. Kestenbaum, J. Dent. Res. 54 (1975) 351. [15] M. Rosenberg, I. Septon, I. Eli, R. Barnes, I. Gelernter, S. Brenner, I. Gabbay, J. Periodontol. 62 (1991) 487. [16] M. Shimura, Y. Yasumo, M. Iwakura, Y. Shimada, S. Sakai, K. Suizuki, S. Sakamoto, J. Periodontol. 67 (1996) 396. [17] E. Maita, in: D. Van Steenberghe, M. Rosenberg (Eds.), Bad Breath: A Multidisciplinary Approach, Leven University Press, 1996. [18] M. Shimura, S. Watanabe, M. Iwakura, Y. Oshikiri, M. Kusumoto, S. Sakamoto, J. Periodontol. 68 (1997) 1182. [19] J. Rodr´ıguez-Fernández, J.M. Costa, R. Pereiro, A. Sanz-Medel, Anal. Chim. Acta 398 (1999) 23. [20] J. Rodr´ıguez-Fernández, R. Pereiro, A. Sanz-Medel, Anal. Chim. Acta 471 (2002) 13. [21] S.R. Bhat, J.M. Eckert, R. Geyger, N.A. Gibson, Anal. Chim. Acta 108 (1979) 293. [22] S.R. Bhat, J.M. Eckert, N.A. Gibson, Anal. Chim. Acta 128 (1981) 263. [23] T. Saito, Talanta 41 (1994) 811. [24] L. Szekeres, Talanta 21 (1974) 1. [25] A. Safavi, A. Rahmani, V.N. Hosseini, Fresenius J. Anal. Chem. 354 (1996) 502. [26] R. Narayanaswamy, F. Sevilla, Analyst 111 (1986) 1085. [27] E.B. Sandell (Ed.), Colorimetric Determination of Trace of Metals, Interscience, New York, 1959. [28] M. Volkan, T. Eroglu, A.E. Eroglu, O.Y. Ataman, H.B. Mark Jr., Talanta 47 (1998) 585. [29] B. Markley, C.E. Meloan, J.L. Lambert, Anal. Lett. 19 (1986) 37. [30] M. Menéndez, Master of Science Report, University of Granada, Spain, 2000. [31] M. Valcarcel, Quality assurance of luminescent qualitative analysis, in: X International Symposium on Luminescence Spectrometry— Detection Techniques in Flowing Streams—Quality Assurance and Applied Analysis, Granada, Spain, June 2002.
A sorbent tube for oral malodour monitoring Julio Rodr´ıguez-Fernández a , Regina López-Fernández a , Rosario Pereiro a , Manuel Menéndez b , José Mar´ıa Tejerina b , Alberto Sicilia b , Alfredo Sanz-Medel a,∗ a
Department of Physical and Analytical Chemistry, University of Oviedo, Julián Claver´ıa 8, 33006 Oviedo, Spain b Department of Estomatology, University of Oviedo, Catedrático José Serrano, s/n 33006 Oviedo, Spain Received 9 May 2003; received in revised form 25 July 2003; accepted 13 August 2003
Abstract Volatile sulphur compounds (VSCs) and particularly hydrogen sulphide are considered as the predominant gases causing oral malodour. In this paper, a simple alarm sensor has been developed for VSCs determination in mouth air. The device consists of a glass tube packed with a solid sensing phase. The VSCs react with the sensing phase to produce a change in the colour of the sensor visible with a naked eye. Different “reagents” were investigated to develop the sensing phase (neocuproine + Cu(II), bathocuproine + Cu(II), resazurin, 2,6-dichlorophenolindophenol and lead acetate), finding the neocuproine + Cu(II) as the best for our purposes. Also, different substrates such as Amberlite XAD-4 and XAD-7 and different trademarks of silica gel were tested as solid supports, being selected the silica gel. A device consisting of a glass tube packed with the sensing phase was optimized and tested with halitosis patients as a rapid illness test and the results compared with those obtained with a commercially available instrument, the HalimeterTM , used for the determination of VSCs in mouth air. The results exhibited acceptable agreement between the proposed “qualitative” alarm sensor and a commercially available technique selected as reference, showing the possibility of using this “visual sensor” to control the halitosis and its evolution with an eventual treatment, by the own patient. © 2003 Elsevier B.V. All rights reserved. Keywords: Visual sensor; Volatile sulphur compounds; Halitosis; Neocuproine–Cu(I)
1. Introduction Halitosis is a general term used to describe unpleasant odour coming out from the mouth air and breath, independently from the source where the odour substances are coming from [1–5]. When the sources are oral it is termed “oral halitosis”, being the most frequent source of breath malodour in healthy subjects. Oral pathogenic halitosis is a common problem affecting the adult population [6,7]. It has been previously reported that volatile sulphur compounds (VSCs), such as methyl mercaptan and, in particular, hydrogen sulphide are the major gases associated with oral malodour [8,9]. These odours arise from microbial degradation from proteins (especially these that contain cysteine and methionine), peptides and aminoacids that are present in saliva, gingival crevicular fluid or in food that is retained on the teeth [10]. ∗ Corresponding author. Tel.: +34-98-510-3474; fax: +34-98-510-3125. E-mail address: [email protected] (A. Sanz-Medel).
0039-9140/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2003.08.014
In some studies [11,12], the relationship between the amount of volatile sulphur compounds and the severity of periodontal disease, or oral prophylactic conditions related to the accumulation of tongue coating, gingival fluid, and saliva putrefaction, has indicated that septic and unsanitary oral conditions are often the major cause of apparent halitosis. For the assessment of oral malodour, two types of methods are generally used: one is the subjective organoleptic examination [13] and the other is the quantitative measurement of VSCs; e.g. using gas chromatography [6,14]. Subjective organoleptic examination by nasal sniffing is the simplest and commonly used approach to directly evaluate expelled mouth air. Although this method has been considered as a reference standard of oral malodour measurement, it raises several problems such as the variation between judges on the ranking of the same sample [15]. On the other hand, quantitative measurement by gas chromatography offers consistent results, but such measurements show several practical problems to be used by the patient or the clinician itself (e.g. too complex system, technical expertise
422
J. Rodr´ıguez-Fern´andez et al. / Talanta 62 (2004) 421–426
required, and procedures are costly and time-consuming [6,11]). Due to the growing interest of this topic in the medical field, new techniques for measurement of halitosis related sulphides have been reported in recent years [16–18]. In most cases, these techniques are based on redox measurements and include a zinc-dioxide semiconductor sensor [16], a tin-oxide semiconductor gas sensor [17] (the BB ChekerTM , Tokkuyala Corp., Tokyo, Japan), and the HalimeterTM developed by Interscan Corporation (Chatsworth, CA 91313-2496, US). The last one seems to be the most common instrument for monitoring sulphide-type compounds in the oral cavity. More recently, it has been commercialised an amperometric sulphide silver sensor (Diamond General Development Corp., Michigan, US) which permits the measurement of VSCs in localised regions inside the mouth. To our knowledge only two optical sensors has yet been developed for measurement VSCs in mouth air [19,20]. One is based on the measurement of the fluorescence produced by the reaction of the complex Hg(II)–2,2 pyridylbenzimidazole (immobilised on a silica gel support) with the sulphide ion. The other is a reversible optical fibre sensor based on reflectance measurements of the 2,6-dichlorophenolindophenol on silica gel. All the techniques above mentioned for VSCs monitoring are addressed to the use by the dentist itself, because they require instrumentation more or less sophisticated. Thus, the objective of this work was the development of a simple, cheap and easy to handle “visual sensor” to be used by the patients for the autocontrol of their halitosis in an easy way and without the need of any instrument. For this purpose, different chemicals such as neocuproine + Cu(II), bathocuproine + Cu(II), resazurin, 2,6-dichlorophenolindophenol and lead acetate have been assayed as reagents. Also, different solid supports (Amberlite XAD-4, XAD-7 and silica gel) have been investigated to immobilise the reagents to obtain an adequate “active” phase, changing its colour when VSCs concentration in mouth air surpassed a threshold value. After optimisation for best perception of the colour change, an “active” phase has been selected and evaluated for the detection of halitosis in 72 subjects. The results were compared with those obtained with an electrochemical instrument commercially available (HalimeterTM ) for halitosis monitoring.
drate were purchased from Merck (Darmstadt, Germany). Methanol used to dissolve reagents, was purchased from Prolabo S.A. (Barcelona, Spain). Gallocyanine, toluidine blue O and Congo red, employed as colorants, were purchased from Aldrich. Sodium acetate, sodium tetraborate, sodium hydrogencarbonate, tris(hydroxymethyl)aminomethane, acetic acid, hydrochloric acid, boric acid and sodium hydroxide, all of them from Aldrich, were employed to prepare buffers in the interval from pH 3 to 12. To prepare the neocuproine + Cu(II) and the bathocuproine + Cu(II), the appropriate volumes of organic reagents and copper dissolved in methanol were mixed keeping a molar rate neocuproine: Cu(II), bathocuproine: Cu(II) of 3:1. Amberlite XAD-4, Amberlite XAD-7, silica gel Davisil grade 646 and silica gel Merck grade 10181 purchased from Aldrich and silica gel 40 purchased from Merck were investigated as solid supports. TedlarTM (polyvinyl fluoride) sampling bags with polypropylene valves (Supelco, Madrid, Spain) were used for analyte gas dilution and also for testing the different sensing phases prior to use with patients. Hydrogen sulphide and methyl mercaptan gases were selected as “model” VSCs. They were purchased from Carburos Metalicos (Oviedo, Spain) and were of analytical reagent grade. 2.2. Instrumentation Since visual detection of the analyte was aimed to, no optical instrumentation was used for the sensor. For the preparation of the sensing phase a rotaevaporator Helidoph VV2000 with a water bath Helidoph WB2000 and a pHmeter with a glass electrode (model MicropH 2000, Crison Instruments S.A., Alella, Spain). PyrexTM minicolumns of 8 mm o.d., 3 mm i.d. and 20 mm length were filled with the solid sensing phase (Fig. 1). A common plastic mouthpiece (as those used for tobacco smoking) was inserted at an end of the minicolumn. A 250 ml plastic bag was inserted at the other end. For each test, the patients should blow from the mouthpiece through the minicolumn up to the 250 ml plastic bag is filled,
2. Experimental 2.1. Reagents and solutions All the colorimetric reagents were of analytical grade. Neocuproine hydrochloride hydrate, bathocuproinedisulfonic acid disodium salt hydrate and resazurin were purchased from Sigma–Aldrich (Madrid, Spain), 2,6-dichlorophenolindophenol, lead acetate and copper sulphate pentahy-
N
N
CH3
CH3
2
Cu (I) Fig. 1. Chemical structure of the neocuproine–Cu(I).
J. Rodr´ıguez-Fern´andez et al. / Talanta 62 (2004) 421–426
thus meaning that such air volume has passed through the minicolumn. A commercially available instrument for screening halitosis (HalimeterTM ) was selected for comparison. 2.3. Preparation of the solid sensing phase Methanolic solutions, containing the investigated reagent and the required chemicals as well as the solid support, were evaporated in a rotaevaporator with a water bath at 20 ◦ C (solutions were prepared in methanol). After the rotaevaporation, the solid phase is washed with a convenient solution (when required) and dried at room temperature. Finally, the sensing phase is stored in the refrigerator at 4 ◦ C and in the darkness.
3. Results and discussion 3.1. Selection of the colorimetric reagent For the development of the solid sensing phase different reagents were evaluated: neocuproine + Cu(II) [21,22], bathocuproine + Cu(II) [23], lead acetate [24], resazurin [25], and 2,6-dichlorophenolindophenol [26]. All of them were selected from the literature due to their good analytical performance characteristics for hydrogen sulphide analysis. The solid “active” phases are packed into the glass minicolumn to produce the visual sensor. Then, a known quantity of H2 S gas is introduced into the minicolumn until a change is observed in the colour of the sensing phase and this point is defined as the “visual detection limit” (visual D.L.). Results obtained for the silica gel 40 are shown in Table 1, where it can be seen that the best “visual D.L.” are obtained for neocuproine–Cu and bathocuproine–Cu reagents. Since the analytical behaviour of both chelates proved to be quite similar, being bathocuproine price higher, neocuproine was eventually selected for the further development of an adequate VSCs sensing phase. 3.2. Reagent immobilisation Five different solid supports were compared for the immobilisation of the neocuproine + Cu(II): two nonionic polyTable 1 Selection of the coloured reagent (solid support: silica gel 40) Reagent
Experimental conditions (silica gel concentration)
Visual D.L. (ppb (v/v)) (H2 S)
Neocuproine–Cu(II) Bathocuproine–Cu(II) Resazurin
0.44 g/g 0.055 g/g 0.4 mg/g; Cu(II): 0.16 mg Cu/g 5 mg/g 0.02 g/g
250 250 1000
Lead acetate 2,6-Dichlorophenolindophenol
2500 500
423
Table 2 Selection of the solid support (using the complex neocuproine–Cu(II)) Solid support
Visual D.L. (ppb (v/v)) (H2 S)
Amberlite Amberlite Silica gel, Silica gel, Silica gel,
500 1000 250 250 250
XAD-4 XAD-7 DavisilTM , grade 646 Merck, grade 10181 grade 40
meric adsorbents (Amberlite XAD-4 and Amberlite XAD-7) and three different silica gel supports (Davisil grade 646, Merck grade 10181 and silica gel grade 40). Solid sensing phases were prepared as previously described. Five identical neocuproine + Cu(II) methanolic solutions were prepared by dissolving 0.055 g of the reagent in 50 ml of methanol and adding 10 ml of 7.45 × 10−3 M CuSO4 . Then, to each of the solutions was added 1 g of the corresponding solid support and carried out to dryness into the rotaevaporator at 20 ◦ C. Finally, the solid phases were washed with distilled water up to clear washing liquids and dried at room temperature. Sensing phases prepared in this way were packed into the glass minicolumn and their performance was assessed with hydrogen sulphide samples prepared by dilution of the appropriate H2 S (g) in a TedlarTM bag containing wet air. All silica gel supports (silica gel 40, Davisil grade 646 and Merck grade 10181), offered good visual detection limits according to our purposes (Table 2), but the stability of the reagent in the silica gel Merck grade 10181 and the Davisil grade 646 was worse than in silica gel 40. Therefore, silica gel 40 was selected as the solid support. Similar detection limits were obtained for hydrogen sulphide than for methyl mercaptan. 3.3. Neocuproine + Cu(II) (effect of reagent concentration) The reaction of the neocuproine + Cu(II) with hydrogen sulphide is a redox process where a change in the colour is produced due to the reduction of the complex by the sulphide ion [21,22]. Neocuproine has been traditionally used with the aid of reducing agents for the determination of Cu [27]. Neocuproine + Cu(II) reacts with sulphide ions (Fig. 1) producing a complex with a different colour as a result of the reduction of the Cu(II) to Cu(I) and this reaction can be used for the determination of H2 S [21]. To avoid complex dissociation from the copper bischelate to the monochelate, the molar ratio neocuproine:Cu has to be 3:1. Different concentrations of neocuproine + Cu(II) immobilised on silica gel 40 were tested. It was observed that for concentrations in the interval 0.44–0.055 g of neocuproine (and the Cu(II) in the proper amount for obtaining a 3:1 ratio) per gram of solid support, there was no significant changes in the “visual detection limit” observed. On the other hand, the stability of the sensing phases increased at lower concentrations. Thus, 0.055 g of neocuproine per
424
J. Rodr´ıguez-Fern´andez et al. / Talanta 62 (2004) 421–426
gram of the solid support was eventually selected providing a net colour change for 250 ppb (v/v) of H2 S (g) concentration in air. The stability of this “active” phase was at least 3 months storage (in the refrigerator at 4 ◦ C and so in the darkness). 3.4. pH effect in the preparation of the sensing phase Using 0.055 g neocuproine + 10 ml 7.45×10−3 M CuSO4 per gram of silica gel 40, the optimisation of the pH of the sensing phase was investigated by washing it with different buffers and then, drying it again at room temperature. Different pHs were assayed (3, 4, 5.4, 7.5, 8.5 and 12), but neither the sensitivity nor the stability of the sensing phase were observed to be affected. Therefore, the sensing phase was just washed with distilled water in subsequent experiments. 3.5. Effect of a “screen” dye addition When neocuproine + Cu(II) reacts with hydrogen sulphide in the solid support, the chelate is transformed into neocuproine−Cu(I), changing its colour from slight green to yellow. Since we are using “visual detection”, the colour changes are not always easily perceptible for low concentrations of analyte. For this reason, different “screen” dyes were assayed (in order to change the perceived colours before and after the reaction, making the visual change easier to be detected). These “screen” dyes work like optical filters making the colour changes, produced by the presence of the hydrogen sulphide, more perceptible. The assayed dyes were: resazurin (violet), toluidine blue O (blue), gallocyanine (blue) and Congo red (red). These dyes were added to the mixture before rotaevaporation, and then carried out to dryness following the procedure described in Section 2.3. No enhancement of the perception of the colour changes with single or different mixtures of the above dyes was observed.
3.6. Final design of the sensor The final design of the sensor sought should give rise to a robust and simple device as such visual sensors are going to be used by people who have not received special training for their use. Also, the device should be cheap to help the commercialisation of the product. Attending these reasons, two approaches were assayed: glass minicolumns filled with the sensing phase [28] and paper sticks with the sensing phase on the surface of the paper [29]. After several experiments, the approach of using minicolumns filled with the sensing phase was selected. Different tubes (1.25, 2.0, 3.0, 4.0 mm i.d.) were assayed looking for a good “visual detection limit”. The used tubes were made of PyrexTM with thick walls (8 mm o.d.) to obtain a visual effect similar to a lens. The best analytical results were obtained with the tube with thinner internal diameter. Unfortunately, on using such devices the patients had problems blowing through this thin tube (due to the pressure created inside the tube by the sensing packed material), therefore, a tube with 3.0 mm i.d. was eventually chosen as a compromise between good VSCs detectability and easy blowing through it. The sensing phase proved to be non reversible over a period of 24 h after its use. To optimise the perception of the colour change, the sensing phase was eventually placed between two slices of white silica gel (trying to get a good contrast between the colours) within the minicolumn (see Fig. 2). 3.7. Interference studies The interference studies were carried out by simulating, as much as possible, the behaviour of possible interferent compounds present in the mouth air and breath. Different samples of mouth air were sampled from several volunteers without halitosis problems (control) under different conditions and evaluated with the “visual sensor” developed. No colour changes were observed. The samples were doped with 300 ppb (v/v) of H2 S (g) and different conditions of mouth
Filters
2.5 mm
3 mm
250 mL bag 250 mL bag
White silica gel
Mouthpiece
Sensing phase
Fig. 2. Schematic representation of the sensing system, including the mouthpiece and the plastic bag.
J. Rodr´ıguez-Fern´andez et al. / Talanta 62 (2004) 421–426
air were studied: after smoking a cigarette, after cleaning the tooth with a fluoride tooth paste, after cleaning the tooth with an elixir, after eating chewing gum with chlorophyll and, finally, after drinking alcohol. No difference between the samples with and without the potential interference was noticed. Therefore, such common mouth air environments can be considered not disturb the “visual sensor” performance. 3.8. Analysis of samples from halitosis patients For the validation of the proposed “alarm test”, comparative tests were carried out with halitosis patients and with volunteers without halitosis problems at the Department of Estomatology at the University of Oviedo. In all cases, the “visual sensor” results obtained by the dentists were compared with those obtained in parallel with an alternative quantitative technique (HalimeterTM ) used for malodour monitoring.
425
Comparative studies carried out in close collaboration with the Department of Estomatology from the University of Oviedo/University of Granada, indicated that a “socially accepted” VSCs limit in mouth air (organoleptic tests versus Halimeter) could be taken as about 250 ppb (v/v) [30]. As the “visual” sensor developed is a rather qualitative device (“alarm sensor”) there are only two possible lectures: positive (+) when the level of the VSCs in mouth air is superior to 250 ppb (v/v) (halitosis patient), and negative (−) when the value is lower than 250 ppb (v/v). Table 3 shows a comparison between the two methods of the values obtained for the control volunteers without halitosis problems and patients. As can be seen, a good correlation between the results obtained with our “visual” sensor and those obtained with the HalimeterTM was observed. From the results, a 2.8% of false negatives and a 5.6% of false positives were obtained. The percentage reliability according to Valcarcel [31] (reliability(%) = 100 − false negatives (%) − false positives (%)) was 91.6% (commonly, reliability
Table 3 Mouth air detection Volunteers
HalimeterTM values
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
119 215 119 262 102 190 118 127 156 120 132 128 177 172 96 163 158 169 1034 453 425 757 348 280 232 1658 542 408 1673 220 258 187 677 380 309 341 ∗
Unexpected results.
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
10 17 10 21 8 15 9 10 12 10 11 10 14 14 8 13 13 14 62 27 26 45 21 17 20 99 33 24 100 13 15 11 41 23 19 20
Visual sensor values
Volunteers
HalimeterTM values
− − − − − − − − − − − − − − − − − − + + + + + + + + + + + +∗ + +∗ + + + +
37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
96 261 250 191 121 234 210 110 223 234 220 295 193 208 329 209 154 258 300 928 136 115 491 240 1323 435 1401 886 752 342 252 546 256 260 546 309
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
8 21 20 15 10 19 17 9 18 19 18 24 15 17 26 17 12 21 18 56 8 7 29 14 79 26 84 53 45 21 15 33 15 16 33 19
Visual sensor values − − − − − − − − − − − −∗ − − −∗ − − − + + +∗ +∗ + + + + + + + + + + + + + +
426
J. Rodr´ıguez-Fern´andez et al. / Talanta 62 (2004) 421–426
percentages lower than 80% are considered as unacceptable [31]). Therefore, the results of Table 3 demonstrate the utility of the proposed “visual sensor” as an “alarm test” for the direct control of the halitosis by the own patient. 4. Conclusions An “alarm” visual sensor for oral malodour is proposed. The developed “visual sensor” is based on the colour change of a sensing phase of neocuproine–Cu and this device offers the possibility of daily control of halitosis problems by the patient himself. The proposed device is very simple to use and does not require any sample pre-treatment of mouth air. The non reversibility of the chemical system should not constitute a problem, because the sorbent tube was thought as something cheap of just one use (for example, to include a set of them into any already available commercial kit for bad breath cleaning). Besides, the production of the sensing phase is very easy. In any case, it should be taken into account that although neocuproine could be considered as a hazardous compound, this should not be a problem because it does not get in contact with the potential user. In any case, a quality control must be carried out in an eventual massive routine preparation of the sorbent tubes. Acknowledgements The authors thank Dentaid S.A. (Cerdanyola, Spain) for lending the HalimeterTM for mouth air determinations. Financial support from FICYT (Spain) and Dentaid S.A. (Cerdanyola, Spain) through the cooperative project number PA-SAL-97-01 is gratefully acknowledged. Thanks are particularly given to Mr. Enric Masdevall and Mr. Jaime Arum´ı for their continuous help and interest. Also, financial support from “Plan Nacional de I + D” (Spanish Ministry of Science and Technology) through the project Ref. PPQ2000-1291-C02-02 is gratefully acknowledged. References [1] M. Sanz-Alonso, ROE 1 (1996) 97. [2] M. Rosenberg, C. McCulloch, J. Periodontol. 63 (1992) 776.
[3] K. Yaegaki, J.M. Coil, J. Can. Dent. Assoc. 66 (2000) 257. [4] M. Quirynen, J. Clin. Periodontol. 30 (5) (2003) 17. [5] J. Rodr´ıguez, R. Pereiro, A. Sanz-Medel, Spectrosc. Eur. 14/2 (2002) 6. [6] J. Tonzetich, J. Periodontol. 48 (1977) 13. [7] M. Shimura, S. Watanabe, M. Iwakura, Y. Oshikiri, M. Kusumoto, K. Ikawa, S. Sakamoto, J. Periodontol. 68 (1997) 1182. [8] J. Tonzetich, Archs. Oral Biol. 16 (1971) 587. [9] N.F. Schmidt, S.R. Missan, W.J. Tarbet, A.D. Cooper, Oral Surg. Oral Med. Oral Pathol. 45 (1978) 561. [10] W.J. Loesche, E.H. DeBoer, in: D. Van Steenberghe, M. Rosenberg (Eds.), Bad Breath: A Multidisciplinary Approach, Leven University Press, 1996 (Chapter 3). [11] Y. Yasuno, M. Iwakura, Y. Shimada, J. Dent. Health 39 (1998) 663. [12] H. Miyazaki, C. Fujita, I. Soh, T. Takewara, in: D. Van Steenberghe, M. Rosenberg (Eds.), Bad Breath: A Multidisciplinary Approach, Leven University Press, 1996. [13] M. Rosenberg, G. Kulkany, A. Bosy, C. McCulloch, J. Dent. Res. 70 (1991) 1436. [14] M. Solis-Gaffar, H. Neiles, W.C. Rainieri, R. Kestenbaum, J. Dent. Res. 54 (1975) 351. [15] M. Rosenberg, I. Septon, I. Eli, R. Barnes, I. Gelernter, S. Brenner, I. Gabbay, J. Periodontol. 62 (1991) 487. [16] M. Shimura, Y. Yasumo, M. Iwakura, Y. Shimada, S. Sakai, K. Suizuki, S. Sakamoto, J. Periodontol. 67 (1996) 396. [17] E. Maita, in: D. Van Steenberghe, M. Rosenberg (Eds.), Bad Breath: A Multidisciplinary Approach, Leven University Press, 1996. [18] M. Shimura, S. Watanabe, M. Iwakura, Y. Oshikiri, M. Kusumoto, S. Sakamoto, J. Periodontol. 68 (1997) 1182. [19] J. Rodr´ıguez-Fernández, J.M. Costa, R. Pereiro, A. Sanz-Medel, Anal. Chim. Acta 398 (1999) 23. [20] J. Rodr´ıguez-Fernández, R. Pereiro, A. Sanz-Medel, Anal. Chim. Acta 471 (2002) 13. [21] S.R. Bhat, J.M. Eckert, R. Geyger, N.A. Gibson, Anal. Chim. Acta 108 (1979) 293. [22] S.R. Bhat, J.M. Eckert, N.A. Gibson, Anal. Chim. Acta 128 (1981) 263. [23] T. Saito, Talanta 41 (1994) 811. [24] L. Szekeres, Talanta 21 (1974) 1. [25] A. Safavi, A. Rahmani, V.N. Hosseini, Fresenius J. Anal. Chem. 354 (1996) 502. [26] R. Narayanaswamy, F. Sevilla, Analyst 111 (1986) 1085. [27] E.B. Sandell (Ed.), Colorimetric Determination of Trace of Metals, Interscience, New York, 1959. [28] M. Volkan, T. Eroglu, A.E. Eroglu, O.Y. Ataman, H.B. Mark Jr., Talanta 47 (1998) 585. [29] B. Markley, C.E. Meloan, J.L. Lambert, Anal. Lett. 19 (1986) 37. [30] M. Menéndez, Master of Science Report, University of Granada, Spain, 2000. [31] M. Valcarcel, Quality assurance of luminescent qualitative analysis, in: X International Symposium on Luminescence Spectrometry— Detection Techniques in Flowing Streams—Quality Assurance and Applied Analysis, Granada, Spain, June 2002.