Teófilo R. F., Ceragioli H. J., Peterlevitz A. C., Baranauskas V., Ferreira M. M. C., Kubota L. T., "Simultaneous determination of guaicol and chloroguaiacol by SWV using boron-doped diamond electrode and PLS algorithms". Águas de Lindóia, SP, Brazil, 10-15/09/2006: 10th International Conference on Chemometrics in Analytical Chemistry (CAC-2006, CAC-X), Book of Abstracts (2006) P009. Poster 009.

10th International Conference on Chemometrics in Analytical Chemistry P009

Simultaneous determination of guaiacol and chloroguaiacol by SWV
using boron-doped diamond electrode and PLS algorithms

Reinaldo F. Teófilo¹*, Helde J. Ceragioli², Alfredo C. Peterlevitz², Vitor Baranauskas²,
Márcia M. C. Ferreira¹, Lauro T. Kubota¹  teofilo@iqm.unicamp.br

1-Instituto de Química, Universidade Estadual de Campinas
2-Faculdade de Engenharia Elétrica, Universidade Estadual de Campinas

Keywords: guiacol, chloroguaiacol, SIMPLS, PPLS, boron-doped diamond
______________________________________________________________________________________
 

      Contamination  of  water  resources   or    waste  sites    with   organic  pollutants   represents  a   serious
environmental problem.  The determination  of phenol and  its  derivative compounds is  of  great  importance,
since these species are released into the environment  by  a large number of industries.  In addition,  some of
these compounds,  such as  guaiacol and chloroguaiacol,  have been determined  to  be recalcitrant,  toxic to
aquatic species, genotoxic,  lipophilic with propensity  for bioaccumulation.  There  is considerable  interest in
their measurement  in  environmental matrices¹.   Analytical methods  for  the  detection  and quantification of
mixtures of phenols  are  usually  based  on  analytical separation techniques,  which  allow  the identification
and quantification of individual constituents.   Many methods have been developed  for  the  determination  of
phenolic compounds,  such  as  chromatographic,  fluorimetric  and  spectrophotometric  methods.   However,
these techniques do  not  easily  allow  continuous  monitoring,  they  are  expensive,  time-consuming,  need
skilled  operators.   Thus,   the development  of  new methods,  that  allows  the  simultaneous  determination,
without  previous separation  of  these  compounds  is  a  relevant  subject of research.   However,  very  few
reports have described  the employment of  electrochemical techniques for simultaneous phenols  detection².
      The aim of this work was presenting a method to determine simultaneously guaiacol  and  chloroguaiacol
by  Square Wave  Voltammetry  (SWV)  using  boron-doped  diamond  electrode  (BDDE)  and  Partial  Least
Squares (PLS) regression.  Among the electroanalytical techniques used nowadays,  SWV  has proved to be
extremely sensitive  for the detection  of organic molecules.   BDD  electrodes  have received much attention
recently  due to a very large electrochemical window resulting  from  the  low reactivity of their surface.   PLS
was  used  owing to necessity  of  quantification  of  guaiacol and chloroguaiacol  that  presents peaks highly
overlapped.
      Boron-doped electrode was grown  and characterized  by  our research group.   SWV  experiments  were
performed  using  an  Autolab  potentiostat  (PGSTAT20).   Pt  wire  was  used  as  counter electrode  and  a
saturated calomel electrode  (SCE)  as reference.   The potential was scanned  in the range  0.5  up to 1.2 V
and the operational conditions optimized were  35 Hz,  0.002 V  and  0.08 V  to the frequency,  step potential
and amplitude, respectively.  The voltammograms obtained were submitted to baseline correction by  moving
average with peak width of  0.01.   The concentration range to  guiacol and chloroguiacol was  from 2.0x10^-6
to 3.0x10^-5 mol L^-1.   The measurements were carried out in  5 ml of  buffer  McIlvaine  pH 3.25,  0.05 mol L^-1
containing  the  mentioned analytes.  A cathodic treatment in -3 V during 3 s,  under  vigorous agitation,  after
at least 5 sequential analyses,  was performed  to  keep  the precions of measurements.  Two  different  PLS
algorithms were applied,  the SIMPLS³  and  the powered PLS  (PPLS)⁴.  The variables were  meancentered
and selected. The number of samples in the calibration/prediction sets were of 20/10  and 19/10 for guaiacol
and chloroguaiacol,  respectively.   The number  of  factors in the model was determined  by  cross-validation
applying the leave-one-out method,  based  on the calculation of root mean square errors of  cross-validation
(RMSECV).   The  parameters  presented  i n  the  table  indicate  the  model  predictive  ability  for  unknown
samples using  two  different algorithms.   The parameters considered  for an external validation set were the
root mean square errors (RMSEP),  the  correlation  coefficient  (r)  and  the  relative  error.   Three  principal
conclusions  can  be  obtained  according  to  the  results  presented  on  the  table  below:   (i)  the  guaiacol
prediction was significantly better  comparing  to  the chloroguaiacol;   (ii)  the simultaneous  determination of
these compounds using  BDDE  is possible even in low concentrations  despite  the high voltammetric signal
overlapping;  (iii)  the PPLS algorithm showed to be slightly better than SIMPLS to these data sets.

Acknowledgment: The authors thank CNPq for financial support.
______________________________________________________________________________________
______________________________________________________________________________________
___________RMSECV___Factors_____RMSEP________r___________Relative error (%)_____________
              _____________________________________________________A_______________B________
              ___A______B_____A__B____A_____B_____A____B___min__mean__max__min__mean__max__

Gua             0.0085    0.0064         4      4      0.0035   0.0028      0.99     0.99     0.27      1.81      5.06     0.03      1.30      3.89
Cgua           0.0116    0.0115         6     5      0.0160    0.0070      0.97     0.99    3.63    14.40    33.37     1.70    11.20    24.20
______________________________________________________________________________________
A - SIMPLS, B - PPLS

References

¹ Freire R. S.; Thongngamdee S.; Duran N.; Wang J.; Kubota L. T. Analyst 2002, 127, 258-261.
² Freire R. S.; Ferreira M.; Duran N.;  Kubota L. T. Anal. Chim. Acta 2003, 485, 263-269.
³ de Jong S. ;Chemom. Intell. Lab. Syst. 1993, 18, 251-263.
⁴ Indahl U.; J. Chemometr. 2005, 19, 32-44.