Anal Bioanal Chem (2005) 383: 864–868 DOI 10.1007/s00216-005-0055-0
ORIGINA L PA PER
Peter Tomčík . Pavlína Jenčušová . Monika Krajčíková . Dušan Bustin . Roman Brescher
The detection of formaldehyde in textiles using interdigitated microelectrode array diffusion layer titration with electrogenerated hypobromite Received: 10 June 2005 / Revised: 26 July 2005 / Accepted: 27 July 2005 / Published online: 20 September 2005 # Springer-Verlag 2005
Abstract An interdigitated microelectrode array (IDA) was applied to the determination of formaldehyde released from textiles produced in industry. The proposed method is based on formaldehyde reaction with hypobromite which is formed in weakly basic media by control current electrooxidation of bromide on the generator segment of the IDA array. The unreacted hypobromite diffuses through the gap between individually polarisable IDA segments and it is amperometrically detected on the collector segment of the IDA. The efficiency of this nonconvective transfer process in the absence of formaldehyde was substantially higher (78%) in comparison with that when using the rotating ring disc electrode. The influence of the added formaldehyde on the transfer process can be utilised to develop a simple and sensitive analytical procedure for formaldehyde detection with a detection limit of 4×10−6 mol dm−3. Keywords Formaldehyde . Interdigitated microelectrode array . Hypobromite . Environmental pollutants
Introduction Formaldehyde is an environmentally important toxic species. It is very often used for various applications such as manufacturing building plates, plywood, and lacquer materials from formaldehyde resins [1, 2]. Another possible source of contamination is environmental technology where formaldehyde can be released from hexamethylenetetraP. Tomčík (*) . P. Jenčušová . M. Krajčíková . D. Bustin Department of Analytical Chemistry, Slovak University of Technology, Radlinského 9, 812 37 Bratislava, Slovak Republic e-mail:
[email protected] R. Brescher Analytical Chemistry Laboratory, VIPOTEST s.r.o. Partizánske, Service Púchov, T.Vansovej 1054, 020 01 Púchov, Slovak Republic
amine used in wastewater purification processes [3]. As a small very reactive molecule, it is widely used in chemical synthesis, for example, as an intermediate in the production of detergents and shampoos. The toxicity for some microorganisms is utilised when formaldehyde is used as a sterilising agent in medicine. Natural concentrations of formaldehyde can be found in living organisms [4] and fruits or vegetables [5]. Because of this, simple and sensitive analytical methods are needed for its detection/ determination. Gas chromatography and high-performance liquid chromatography [6–10] are widely used for detection of formaldehyde. They involve expensive equipment, and a further derivatisation step to allow detection, which could be timeconsuming. The detection limits are in the milligrams per liter range. Chemiluminiscence [11] is sensitive with a detection limit in the nanograms per liter range but it also needs a reaction for the creation of a suitable luminophore. Spectrophotometry with a detection limit of several tens of micrograms per liter [12, 13] is the commonest method for the detection of formaldehyde. Various biosensors [14–20] sensitive to formaldehyde have also been developed. They belong to powerful tools in analytical chemistry, they are very specific, but there are some problems with the stability of enzyme layer, they need to be stored at low temperature and their time response is quite long as well. Therefore we suggest a novel, simple, fast and sensitive method for the determination of formaldehyde in textiles based on interdigitated microelectrode arrays (IDA) fabricated microlitographically [21, 22] which are very promising tools in electroanalysis. They contain individually polarisable segments allowing the generation of one species on the first segment (generator) and its detection on the second segment (collector). This dual voltammetric mode is a principle of very sensitive and selective determination when the IDA response is enhanced by redox cycling [23, 24]. The mixed galvanostatic–voltammetric dual mode was used for the IDA diffusion layer titration of formaldehyde where the titrating agent (hypobromite) is galvanostatically (applying a galvanostatic scan) generated in situ from a precursor present in a supporting electrolyte. Analytical
865
techniques for dithiocarbamates [25, 26] or pharmaceuticals [27] were developed utilising this approach. The optimisation of experimental conditions [26] and the study of more the complicated titration reaction mechanism of iodide with electrogenerated hypobromite were also performed [30]. In our recent paper [31] we presented a microanalytical method for the determination of formaldehyde in wastewater based on its bulk phase titration with hydroxylamine where an IDA serves as a biamperometric detector of the end-point. This method has low sensitivity (millimoles per liter range) and the titration agent should be standardised before each analysis.
Experimental Interdigitated microelectrode array A thin film based on planar horizontally separated a IDA was used (Fig. 1). It contained two individually polarisable comb-shaped segments with a Pt film deposited by rf sputtering and patterned with a photolitographical lift-off technique [28]. Each segment contained 26 fingers of length 0.9 mm and width 5 μm. The gap between adjacent fingers of the two segments was 5 μm. Construction of the IDA was accomplished on alumina–boron glass with dimensions 15 mm×3 mm×0.6 mm, and was improved with a polyimide film to obtain enhanced stability in media above pH 7.
The pH was checked with a model 215 pH meter (Denver Instruments, USA) with a glass electrode. The reference electrode was a standard calomel electrode (SCE) and the counter electrode was a carbon rod of area 1.5 cm2. Chemicals and procedures Hypobromite was generated from 1 M NaBr (Lachema CZE) with added 0.1 M sodium tetraborate (Lachema CZE). The pH was adjusted with dilute H2SO4 (Lachema CZE) to the desired value. A stock standard solution of formaldehyde was standardised iodometrically. Two grams of textile was weighed and then placed in a beaker with 30 ml of triply distilled water. The beaker was closed and placed in a water bath for 60 min at 40°C. After cooling and filtering. the final solution was transferred to a 50-ml volumetric flask.
Results and discussion Preliminary study of the generating–collecting process To understand in detail how the generating–collecting transfer proceeds the dependencies of the generator and the collector current were registered. The generator was operated in galvanostatic mode, producing hypobromite from NaBr in sodium tetraborate at a pH value corresponding to weakly alkaline media according to the electrode reaction
Equipment
Br þ 2OH ! BrO þ 2e þ H 2 O:
Electrochemical measurements were carried out on an EG&G Princeton Applied Research model 366 bipotentiostat which allows independent polarisation of two working electrodes and galvanostatic sweeping one of them.
The generator current was changed in time according to the equation I ¼ I0 þ vt;
(1)
(2)
where I0 is starting current (normally equal to zero), I and t are actual values of the generator current and time, respectively, and v represents how fast the generator current changes with the current scan rate dI/dt in nanoamperes per second. The charge passed during electrogeneration of the titrant in time t can be expressed as an integral of Eq. (2): Q ¼ 0:5t 2
(3)
and number of moles of titrant n¼ 0:5t 2 =zF:
(4)
The time flux of the titrant is the first derivate of Eq. (4): dn=dt¼t=zF: Fig. 1 Schematic geometrical depiction of the horizontally separated platinum interdigitated microelectrode array used (not to scale), Wgap=5 μm, We=5 μm
(5)
866
Fig. 2 Consecutive runs of generator vs. collector current dependence without addition of formaldehyde, current scan rate 5 nA s−1
The collector served as an amperometric detector of hypobromite arriving from the generator. The potential of the collector was kept constant at of 0.0 V vs. the SCE, corresponding to the limiting current of the reaction opposite to that in Eq. (1). The value was determined from the cyclic voltammogram (not shown). The generator vs. collector current dependencies (without addition of formaldehyde) exhibit a small intercept called a blank [29]. This blank is different for each IDA of this type. It is a constant and highly reproducible parameter for a given IDA and is probably connected with surface processes on the IDA. This should be subtracted when doing model and real sample analysis. In Fig. 2 several consecutive generator vs. collector current dependencies are shown. The value of the blank was calculated to be 70±1 nA (six runs) for a current scan rate of 5 nA s−1 . The blank also depends on the current scan rate. The generator vs. collector current dependencies for various current scan rates in the range 5–300 nA s−1are depicted in Fig. 3. From this dependence it can be deduced that the blank increases
Fig. 4 Dependence of the time response (defined as blank-togenerator current scan rate ratio) of the collector on the generator current scan rate
with increasing generator current scan rate. The time response of titrant transfer defined as the time at which the collector current starts to grow owing to hypobromite detection decreases exponentially with current scan rate (Fig. 4) owing to the large flux of electrogenerated species (Eq. 5) arriving on the detector. Owing to its nonlinear decrease the blank will increase with current scan rate. In addition, the collection efficiency defined as the slope of this dependence is decreased when the generator vs. collector current dependence is registered quickly. For slower scans the value of the collection efficiency was 0.78. This value is in good agreement with values obtained in our previous experiments [29] and is substantially higher than in the case of a rotating ring-disc electrode, because the species is transferred to the collector by diffusion instead of convection. For good precision and accuracy of blank determination point-by-point construction of generator vs. collector current dependencies of the titration curve is optimal but it takes a long time; therefore, we strongly recommend slow scans from 5 to 10 nA s−1.
Diffusion layer titration of formaldehyde The transfer of hypobromite from the generator to the collector of the IDA is influenced by addition of formaldehyde. Formaldehyde is electroinactive at the potential of hypobromite generation as well as at the collector potential of 0 V vs. the SCE. Its reaction with hypobromite has two steps Fig. 3 Generator vs. collector current dependencies without addition of formaldehyde for different current scans rates: 5, 10, 20, 50, 100, 200, 300 nA. s−1 (from the left to the right)
BrO þHCHO ! HCOOH þ Br
(6a)
867 Table 2 Analysis of real samples (textile pieces) of formaldehyde; the number of analyses was 6 Sample
A B C D
Formaldehyde SD content (mg kg−1) −1 a (mg kg ) 207 240 353 400
13 15 18 23
Formaldehyde SD (mg kg−1) content −1 b (mg kg ) 193 231 367 390
9 13 21 24
a
Determined by diffusion layer titration on an interdigitated microelectrode array b Determined by independent analysis in the Vipotest laboratory
Fig. 5 Diffusion layer titration curves for various concentrations of formaldehyde: 1 0 mol dm−3, 2 1.1×10−4 mol dm−3, 3 2.2×10−4 mol dm−3, 4 3.3×10−4 mol dm−3, 5 4.4×10−4 mol dm−3; the blank was subtracted
BrO þHCOOH ! CO2 þBr þ H2 O
(6b)
and the overall reaction is
2BrO þHCHO ! 2Br þCO2 þH2 O:
(6c)
This reaction is fast and quantitative in 1 M NaBr with addition of sodium tetraborate at pH 7.7 which is also optimal for hypobromite electrogeneration. It only takes place in the vicinity (diffusion layer) of the generator electrode (diffusion layer titration) since BrO- is not present in the bulk of the solution analysed. The diffusion layer titration curve (generator vs. collector current dependence) has a different shape (larger intercept) in comparison with that in the absence of formaldehyde (Fig. 5). Electrogenerated hypobromite reacts with formaldehyde and when the flux of hypobromite is lower than the flux of formaldehyde the collector current remains zero. When the flux of hypobromite is larger it reaches the collector segment and the collector current Table 1 Analysis of model samples of formaldehyde (a base electrolyte 1 M NaBr with sodium tetraborate pH 7.7, volume 10 ml); the number of analyses was 10 Taken Found Confidence interval for 95% SD (mg kg−1) (mg kg−1)a (mg kg−1) probability (%)b 5.6 10.2 16.3 19.4
5.2 9.9 15.1 19.9
0.5 0.8 1.1 1.3
0.36 0.57 0.79 0.93
6.9 5.8 5.2 4.7
SD standard deviation a Average value from ten measurements b Calculated according to the formula tn−1, αSD/sqrt(n); t9; 0.05=2.2622
starts to increase. By extrapolation of the linear portion of the collector current to the collector residual current, the generator current at the titration equivalent point IgenE was obtained. This parameter is linearly dependent on the concentration of formaldehyde in bulk-phase solution and can be utilised for its quantification. The relation between IgenE and the concentration of formaldehyde can be expressed by the equation IgenE ¼ intercept þ slope c formaldehyde ;
(7)
where the intercept is 8.1 nA, the relative standard deviation (RSD) is 12%, and the slope is 659,410 nA mol−1 dm3, the RSD is 1% (current scan rate 5 nA s−1). The estimate of the detection limit of this method according to the criterion 3SDintercept =slope
(8)
is 4×10−6 mol dm−3 (120 μg kg−1). It represents in our experimental conditions 1.2 μg of formaldehyde (absolute amount). The results of the analysis of four samples with 5–20 mg kg−1 formaldehyde are summarized in Table 1. Statistically the mean does not differ from the known value. This makes this method suitable for analysis of real samples . Textile pieces of unknown origin served as real samples. The results of the analyses of real samples are given in Table 2 and they are in good agreement with results obtained with a spectrophotometric method based on reaction with chromotropic acid in sulphuric acid (absorbance at 570 nm is measured) as an independent analysis performed in the Vipotest laboratory.
Conclusions An IDA prepared microlitographically was used in the development of a method for the determination of formaldehyde based on its diffusion layer titration with electrogenerated hypobromite. The technique is simple, sensitive and suitable for analysis of simple samples of formaldehyde. It is based on reaction of formaldehyde with electrogenerated hypobromite in the diffusion layer of the IDA
868
and the titration curves can be obtained many times with the same sample solution. The selectivity of this method is given by the selectivity of the titration reaction; therefore, possible other species reacting with hypobromite should be removed or masked because of incorrect determination of IgenE. However this technique is well suited for the detection of formaldehyde in textiles because no other species reacting with hypobromite are present. Acknowledgements This study was supported by the Slovak Grant Agency of Sciences, VEGA (project no. 1/2464/05)
References 1. Gerberich HR, Seaman GC (1994) Formaldehyde: encyclopedia of chemical technology, Vol. 11, 4th edn. Wiley, New York, p 929 2. Patnaik P (1997) Handbook of environmental analysis: chemical pollutants in air, water and solid wastes CRC, Boca Raton, FL 3. Eiroa M, Kennes C, Veiga MC (2004) Water Res 38:3495– 3502 4. Lengyel J, Kalaisz H, Szarvas I, Peltz C, Szarkane-Bolehovszky A (2003) J Chromatogr Sci 41:177–181 5. Flyvholm MA, Andersen P (1993) Am J Int Med 24:533–552 6. Dumas T (1982) J Chromatogr A 247:289–295 7. Luo W, Li H, Zhang Y, Catharina YWA (2001) J Chromatogr B Biomed Sci Appl 753:253–257 8. Medvedovici A, David F, Sandra P, David V (1999) Anal Lett 32:581–592 9. Mann B, Grayeski ML (1987) J Chromatogr 386:149–158 10. Lorrain JM, Fortune CR, Dellinger B (1981) Anal Chem 53:1302–1305 11. Lu XH, Lu MG, Yin F (1992) Anal Lett 25:2073–2081 12. Zhang Z-Q, Yan H-T, Yue X-F (2004) Microchim Acta 146:259–263
13. Gamiz-Gracia L, de Castro MDL (1999) Analyst 124:1119– 1121 14. Ho MH, Richards RA (1990) Environ Sci Technol 24:201–204 15. Winter B, Cammann K (1989) Fresenius Z Anal Chem 334:720 16. Chan WH, Xie TY (1997) Anal Chim Acta 349:349–357 17. Korpan YI, Gonchar MV, Starodub NF, Shu’ga AA, Sibirny AA, El’skaya AV (1993) Anal Biochem 215:216–222 18. Korpan YI, Soldatkin AP, Gonchar MV, Sibirny AA, Gibson TD, El’skaya AV (1997) J Chem Technol Biotechnol 68:209– 213 19. Rindt K-P, Scholtissek S (1989) In: Schmid RD, Scheller FW (eds) Biosensors: applications in medicine, environmental protection and process control. GBF monographs, Vol. 13, VCH, Weinheim, p 405 20. Hämmerle M, Hall EAH, Cade N, Hodgins D (1996) Biosens Bioelectron 11:239–246 21. Dzyadevych SV, Arkhypova VN, Korpan YI, El’skaya AV, Soldatkin AP, Jaffrezic-Renault N, Martelet C (2001) Anal Chim Acta 445:47–55 22. Horiuchi T, Niwa O, Morita M, Tabei H (1992) Anal Chem 64:3206–3208 23. Bustin D, Mesároš Š, Tomčík P, Rievaj M, Tvarožek V (1995) Anal Chim Acta 305:121–125 24. Tomčík P, Bustin D (2001) Fresenius J Anal Chem 371:562– 564 25. Tomčík P, Bustin D (2000) Collect Czech Chem Commun 65:1029–1034 26. Tomčík P, Krajčíková M, Bustin D, Skačáni I (2001) Electrochem Comm 3:191–194 27. Tomčík P, Krajčíková M, Bustin D (2001) Talanta 55:1065– 1070 28. Tvarožek V, Ti Tien H, Novotný I, Hianik T, Dlugopolský J, Ziegler W, Leitmannová-Ottová A, Jakabovič J, Řeháček V, Uhlár M (1994) Sens Actuators B 19:597–602 29. Bustin D, Jursa S, Tomčík P (1996) Analyst 121:1795–1799 30. Tomčík P, Mesároš Š, Bustin D (1998) Anal Chim Acta 374:283–289 31. Tomčík P, Mrafková L, Bustin D (2003) Microchim Acta 141:69–72
