You are on page 1of 5

Talanta 77 (2009) 1155–1159

Contents lists available at ScienceDirect

Talanta
journal homepage: www.elsevier.com/locate/talanta

A portable, inexpensive and microcontrolled spectrophotometer based


on white LED as light source and CD media as diffraction grid
Germano Veras a,b , Edvan Cirino Silva a,∗ , Wellington Silva Lyra a ,
Sófacles Figueredo Carreiro Soares a , Thiago Brito Guerreiro a ,
Sérgio Ricardo Bezerra Santos c
a
Universidade Federal da Paraíba, Departamento de Química, 58.051-970, João Pessoa, PB, Brazil
b
Universidade Estadual da Paraíba, Departamento de Química, 58.109-753, Campina Grande, PB, Brazil
c
Centro Federal de Educação Tecnológica da Paraíba, 58.015-430, João Pessoa, PB, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: A portable, microcontrolled and low-cost spectrophotometer (MLCS) is proposed. The instrument com-
Received 17 April 2008 bines the use of a compact disc (CD) media as diffraction grid and white light-emitting diode (LED) as
Received in revised form 18 August 2008 radiation source. Moreover, it employs a phototransistor with spectral sensitivity in visible region as
Accepted 19 August 2008
phototransductor, as well as a programmable interrupt controller (PIC) microcontroller as control unit.
Available online 27 August 2008
The proposed instrument was successfully applied to determination of food colorants (tartrazine, sun-
set yellow, brilliant blue and allura red) in five synthetics samples and Fe2+ in six samples of restorative
Keywords:
oral solutions. For comparison purpose, two commercial spectrophotometers (HP and Micronal) were
White LED
CD media
employed. The application of the t-paired test at the 95% confidence level revealed that there are not sig-
Programmable interrupt controller (PIC) nificant differences between the concentration values estimated by the three instruments. Furthermore,
microcontroller a good precision in the analyte concentrations was obtained by using MLCS. The overall relative standard
Visible spectrophotometer deviation (R.S.D.) of each analyte was smaller than 1.0%. Therefore, the proposed instrument offers an eco-
Food colorants nomically viable alternative for spectrophotometric chemical analysis in small routine, research and/or
Iron (II) teaching laboratories, because its components are inexpensive and of easy acquisition.
Published by Elsevier B.V.

1. Introduction Therefore, it is possible to develop simple spectrophotometers


using a single white LED with an adequate disperser of the
Since the seventies of the last century, when researchers as white light. For this purpose, the literature reports two works
Flaschka et al. [1] developed photometers based on light-emitting only that employ white LED as radiation source [4,5]. Shimazaki
diode (LED), LEDs have been applied on spectrophotometric instru- et al. [4] were pioneers in the development of a white LED-
mentation as radiation sources having relatively small effective based spectrophotometer which was used in determination of
bandwidth [2,3]. However, the increasing demand for chemical Fe2+ in river water by the o-phenantroline method. Li et al. [5]
analyses makes necessary to develop LED-based instruments to described a portable multi-function instrument using different
work in a wide range of wavelengths what requires multiple spectrophotometric techniques. As application, the spectropho-
radiation sources. The approach increases cost and complexity tometric determination of Cr6+ and Al3+ was carried out using
of the optical and electronic system. For example, optical fibers diphenylcarbazide and pyrocatechol violet as chelant and chro-
are usually employed to direct the radiation towards the detector mogenic reagents, respectively.
and a more elaborated hardware is necessary to control the LED Commercial spectrophotometric instrumentation commonly
drift. employs absorption or interference optical filters, prisms or diffrac-
Nowadays, it is easy to acquire white LEDs that substitutes tion grids as radiation dispersers. In most of these instruments, the
the complex optical system obtained when monochromatic LEDs dispersive device is a diffraction grid, an optical component con-
are used as sources of several wavelengths in the visible region. taining series of grooves traced on a glass plate or polished metal. In
this context, a compact disc (CD) media could be used as diffraction
grid due to its grooves. Despite the use of CD media as radiation dis-
∗ Corresponding author. perser [6], no application in spectrophotometric instrumentation
E-mail address: edvan@quimica.ufpb.br (E.C. Silva). has been found.

0039-9140/$ – see front matter. Published by Elsevier B.V.


doi:10.1016/j.talanta.2008.08.014
1156 G. Veras et al. / Talanta 77 (2009) 1155–1159

In the present work, a portable, microcontrolled and low-


cost spectrophotometer (MLCS) having a white LED as radiation
source and a CD media as diffraction grid for measurements
using the visible region was developed. In order to increase the
signal-to-noise ratio in the spectrophotometric measurements, a
phototransistor with maximum spectral sensitivity in the visible
region was used. In the MLCS, pulsed radiation from the white
LED is dispersed by the CD media generating monochromatic
radiation that is focalized in the phototransistor by a stepper
motor. The control of the proposed instrument, the acquisition
and treatment of data are accomplished by an electronic circuit
based on a programmable interrupt controller (PIC) microcon-
troller. Two analytical applications were elected to illustrate the
performance of the MLCS. The first is represented by individual
determinations of four food colorants (tartrazine, brilliant blue, Fig. 1. Schematic diagram illustrating the optical system of the MLCS.
allura red and sunset yellow). The second concerns to the deter-
mination of Fe2+ in restoratives by using the 1,10-phenantroline
method. solution at pH 7.00. Standard solutions of synthetic samples of
each colorant were prepared by the appropriated dilution of the
respective stock solution. The standard solutions were prepared in
2. Experimental authentic triplicates at the concentrations 2.0, 4.0, 6.0, 8.0, 10.0,
12.0, 14.0, 16.0, 18.0, 22.0, and 26.0 mg L−1 .
2.1. Reagents, solutions and samples FeCl3 ·6H2 O was used to prepare stock solutions of Fe3+
1000 mg L−1 . Working solutions (Fe2+ 2.0–10.0 mg L−1 ) were pre-
Food colorants tartrazine (E-102), sunset yellow (E-110), allura pared from the stock solution after reduction of the Fe3+ using
red (E-129), and brilliant blue (E-133) were purchased from ascorbic acid 1.0% (w/v). Six samples of restorative oral solu-
Sigma–Aldrich. Stock solutions of 1000 mg L−1 of each colorant tions (containing Fe2+ ) from different marks were acquired in
were prepared dissolving its necessary amount in phosphate buffer drugstores at João Pessoa city, Brazil. These pharmaceutical for-

Fig. 2. Electronic layout of the MCLS detector. Legend: PT = phototransistor; C = polyethylene capacitors; R = resistors; U1 = operational amplifier; U2 = quadruple operational
amplifier; U3 = quadruple switches. Details in text.
G. Veras et al. / Talanta 77 (2009) 1155–1159 1157

mulations were diluted to fit in the linear range of the analytical optical–mechanical system. Since the slit for the light entrance
curve. on the probe and the sensitive area of the phototransistor have
Recently distilled water was used to prepare all solutions. the same dimensions (ca. 0.3 mm), the spectra registered by the
ocean optics spectrometer represent the profile of the radiation that
reaches the phototransistor. In this process, the spectrometer reg-
2.2. Description of the proposed instrument
isters the reflected spectrum resulting from the radiation diffracted
on the CD media. It was carried out five scans to obtain an average
Fig. 1 illustrates the optical components employed in the MLCS
spectrum for each step of the stepper motor. Average spectra were
and Fig. 2, the electronic layout of the detection unit. This instru-
used to build a calibration curve between the number of steps and
ment was assembled into a box with 30 cm front, 20 cm height and
the wavelength of maximum emission of the spectrum. In this case,
30 cm length. The optical–mechanical system is composed by white
an equation was estimated by using the linear least-squares and the
LED, focus lens to create a small light image (2 mm diameter) on
result is presented below.
the diffraction grid (CD) and stepper motor with gear to move the
diffraction grid and to promote the scan of the spectral region of Step = wavelength × 0.50246 − 214.7837
interest. To increase the spectral resolution, the position of the step-
per motor was adjusted so that the phototransistor can detect the The number of steps and peak wavelengths are linearly cor-
radiation from the second overtone. related in the spectral range 440–640 nm with a correlation
The instrument uses a PIC 16F877 programmable microcon- coefficient of 0.99968. Therefore, practically the entire visible
troller as control unit driven by a stabilized power supply. The spectral region can be exploited for spectrophotometric measure-
control software was elaborated with C language. The microcon- ments by the MLCS. The overall standard deviation at the peak
troller presents a 10-bit multi-channel analog to digital converter wavelengths was 0.30 nm, indicating a good repeatability asso-
(A/D), 8k × 14 words of flash program memory, 256 × 8 bytes of ciated to the wavelengths obtained starting from the number of
EEPROM data memory, 368 × 8 bytes of RAM, three timers and steps.
three ports with eight channels (multiplexed pins) each, for gen- The collected data were also employed to establish the effective
eral purpose, that can be configured for data acquisition. The bandwidths at maximum wavelength for each step and the value
microcontroller allows drastic reduction of the number of elec- 13 nm was estimated in the range 440–640 nm.
tronic components in the spectrophotometer and makes possible The dispersion of the maximum signals measured for each
to introduce or alter functionalities in the instrument without any step was also assessed and the overall relative standard deviation
hardware modification. (R.S.D.) was estimated as 2.6%. This result indicates a good repeata-
The white LED Control Module pulses the white LED radiation bility of the signals associated to the radiant power of the white
at 7 Hz. The module turns the white LED on for readings of the ana- LED.
lytical signals and off for determination of instrumental noise. The
signals are detected, filtered and amplified in the Detection Module
3.2. Determination of food colorants in synthetic samples
of the MLCS, showed in Fig. 2. In the layout, five operational ampli-
fiers are used for different purposes: to convert the photocurrent
The performance of the MLCS was initially evaluated by ana-
into voltage (U1), to filter the high frequency noise (U2:A) as a high-
lyzing food colorants in the synthetic samples. The absorbance
pass Butterworth filter; to amplify the analytical signal (U2:B), to
measurements were accomplished at the wavelengths of its
detect the peak signal as a sample-hold unit (U2:C) and to fit the
maximum absorption: 486 nm (sunset yellow), 500 nm (allura
analytical signal (U2:D) in the A/D scale (5 V) using CMOS switches
red) and 582 nm (brilliant blue). However, in the case of tar-
(U3:B–D) to change the gain depending on the value of the input sig-
trazine solutions, the measurements were performed at 440 nm
nal. C5 and R11 are used as low-pass filter and P1 is used for off-set.
due to the low sensibility of the MLCS in the wavelength of
CMOS switch U3:A is used to discharge the capacitor of the sample-
maximum absorption of this colorant (426 nm). For reference
hold circuit. The switches are controlled by the PIC. The processed
measurements, the food colorants were also analyzed by an HP
signal E is sent to the A/D converter of the PIC to be converted into
diode-array, model 8453, and a Micronal spectrophotometer, model
absorbance values. To calculate the absorbance values, the loga-
B34211.
rithm function used the measurement of blank (water) as reference.
In order to investigate the maximum concentration for linear
Both blank and sample measurements were registered turning the
response (the limit of linearity—LOL), ANOVA (analysis of variance)
LED on to consider maximum light power and off in order to con-
and F-test for lack of fit were applied to the models elaborated
sider the dark signal. After digitalization, the absorbance values are
with the concentration range 2.0–26.0 mg L−1 , for each colorant.
sent by the microcontroller to the liquid crystal display (LCD) mod-
The results showed no lack of fit up to the LOL as presented in
ule, a TECH2004D-FL-GBS-S character type LCD with two buses:
Table 1. An F-test for the regression significance revealed that the
eight bits for data and three bits for control. In this work, the com-
performed linear regressions are highly significant up to the LOL
munication between microcontroller and LCD was always carried
values.
out with six bits, four being used for data transmission and two
The values of limits of detection (LOD) and limits of quan-
for data control. The Stepper Motor Control Module controls the
tification (LOQ), which were estimated according to IUPAC
angular displacement of the CD media.
recommendations [7], are also shown in Table 1. It is worth not-
A clock elaborated with a 4-MHz piezoelectric crystal is used for
ing that LOL, LOD and LOQ present similar values for the three
synchronization of the microcontroller internal functions.
instruments and for the four food colorants.
After the study above, the analytical curves were constructed
3. Results and discussion employing the following concentration ranges: 2.0–12.0 mg L−1 for
allura red and 3.0–18.0 mg L−1 for the other colorants. In Table 2 are
3.1. Calibration of the MLCS optical–mechanical system presented the results for the colorants determinations in synthetic
samples.
An ocean optics spectrometer, model USB450, which substituted As can be seen, there is a good agreement between the results of
the phototransistor as detector, was used to calibrate the MLCS concentration determined by the proposed instrument and those
1158 G. Veras et al. / Talanta 77 (2009) 1155–1159

Table 1 Tabela 4
Values of LOD, LOQ and LOL obtained for the four colorants Recovery percentage obtained for Fe2+ in six samples of restorative oral solutions

Colorant Instrument LOD LOQ LOL Sample Added Fe2+ (mg L−1 ) Recovery (%)
(10−2 mg L−1 ) (10−2 mg L−1 ) (mg L−1 )
Micronal HP MLCS
Micronal 3.1 9.4 14.0
3.0 99.31 100.34 97.18
Allura red HP 0.7 2.2 14.0
01 5.0 98.86 100.19 100.56
MLCS 0.1 0.2 12.0
7.0 98.98 97.99 96.41
Micronal 3.2 9.7 18.0
3.0 99.65 100.00 97.18
Sunset yellow HP 0.7 2.1 18.0
02 5.0 98.67 99.24 98.05
MLCS 13.0 39.4 18.0
7.0 96.81 97.27 97.25
Micronal 2.6 7.8 18.0
3.0 96.19 94.56 93.85
Tartrazine HP 0.8 3.6 18.0
03 5.0 96.01 98.48 99.86
MLCS 8.1 24.7 18.0
7.0 98.55 99.28 99.05
Micronal 5.5 16.6 18.0
3.0 101.39 98.30 98.72
Brilliant blue HP 7.8 23.8 18.0
04 5.0 99.62 100.19 105.29
MLCS 3.1 9.3 18.0
7.0 99.27 98.28 98.10

3.0 104.51 103.74 97.69


05 5.0 97.91 99.05 97.50
Table 2 7.0 99.42 101.72 100.84
The average concentration values of food colorants and confidence intervals
obtained in the analyses of synthetic samples 3.0 104.51 95.58 103.85
06 5.0 97.91 97.92 94.16
Colorant Sample Expected Estimated value (mg L−1 ) 7.0 99.42 99.43 97.68
value (mg L−1 )
Micronal HP MLCS

01 3.0 2.9 ± 0.1 2.9 ± 0.2 2.9 ± 0.1


02 5.0 4.7 ± 0.0 4.8 ± 0.1 4.7 ± 0.2 were smaller than 1.0%. Such finding indicates a good precision in
Allura Red 03 7.0 6.9 ± 0.1 7.0 ± 0.2 7.0 ± 0.3 the estimative of the colorants concentrations by using the three
04 9.0 8.9 ± 0.0 8.7 ± 0.1 8.9 ± 0.1 instruments.
05 11.0 11.0 ± 0.1 11.0 ± 0.2 11.0 ± 0.4

01 4.0 4.0 ± 0.0 4.1 ± 0.1 4.1 ± 0.2


02 8.0 7.8 ± 0.1 7.8 ± 0.1 7.9 ± 0.2 3.3. Determination of Fe2+ in restoratives oral solutions
Sunset Yellow 03 10.0 10.1 ± 0.1 10.0 ± 0.1 10.0 ± 0.3
04 14.0 13.9 ± 0.1 13.9 ± 0.1 13.8 ± 0.5 In these analyses, the measurements of absorbance were carried
05 16.0 15.9 ± 0.1 15.8 ± 0.1 15.6 ± 0.2
out at 515 nm in the three instruments. Restorative oral samples
01 4.0 4.0 ± 0.1 3.9 ± 0.1 4.1 ± 0.1 of six different manufacturers were analyzed by using analytical
02 8.0 8.0 ± 0.1 7.9 ± 0.1 7.7 ± 0.1 curves built in the concentration ranges 2.0–10.0 mg L−1 , which
Tartrazine 03 10.0 10.2 ± 0.5 10.2 ± 0.1 10.2 ± 0.2
were validated by using ANOVA and F-test for lack of fit and for
04 14.0 13.9 ± 0.5 13.9 ± 0.1 14.0 ± 0.6
05 16.0 15.7 ± 0.0 15.8 ± 0.0 15.8 ± 0.1 regression significance at the 95% confidence level.
Table 3 presents the results in terms of average values of Fe2+
01 4.0 4.1 ± 0.1 4.3 ± 0.1 4.2 ± 0.0
02 8.0 8.0 ± 0.0 7.8 ± 0.1 8.0 ± 0.0
concentration and the confidence intervals estimated at 95% level.
Brilliant Blue 03 10.0 9.9 ± 0.0 9.9 ± 0.1 10.1 ± 0.1 The results show that good agreements of the results were obtained
04 14.0 13.9 ± 0.1 13.7 ± 0.1 13.8 ± 0.1 employing the three instruments and it is corroborated by the
05 16.0 15.9 ± 0.1 15.8 ± 0.1 15.9 ± 0.1 application of the t-paired test at the 95% confidence.
Table 3 also presents the values of overall R.S.Ds., which indicate
a good precision in the estimation of the Fe2+ concentrations by the
obtained by using the commercial spectrophotometers. In fact, the three instruments.
application of the t-paired test at the 95% confidence level revealed In spite of the t-test to reveal absence of systematic error (bias)
that there are not significant differences between the concentra- it does not guarantee its actual absence in the results. In order
tions estimated by the three instruments. to check the eventual presence of bias, a recovery test was car-
Concerning to the precision of results, in Table 2 is shown that ried out according to the procedure described elsewhere [8]. For
the confidence intervals are narrow and similar for the concen- this purpose, known quantities of the analyte were added into the
tration values obtained by the three instruments. Moreover, the restorative samples and the measured recovery taxes are presented
overall R.S.Ds. associated to the results from the three instruments in Table 4. As can be seen, satisfactory recovery taxes were obtained
for all fortified samples. Therefore, this result evidences the absence
Tabela 3 of bias associated to the estimated values of Fe2+ concentration
Average values of Fe2+ concentration and confidence intervals obtained in the anal- employing the three instruments.
yses of samples of restorative oral solutions

Sample Estimated values (mg L−1 ) 4. Conclusions


Micronal HP MLCS

01 4.9 ± 0.1 4.9 ± 0.0 4.9 ± 0.1 A portable, inexpensive and microcontrolled spectrophotome-
02 5.8 ± 0.1 5.7 ± 0.0 5.8 ± 0.1 ter was developed in this work. Such advantageous characteristics
03 4.2 ± 0.1 4.1 ± 0.1 4.3 ± 0.1 were achieved due to the incorporation of a CD media as diffraction
04 6.3 ± 0.1 6.3 ± 0.1 6.2 ± 0.1
grid and a white LED as radiation source. Other relevant char-
05 4.5 ± 0.0 4.4 ± 0.0 4.5 ± 0.0
06 9.0 ± 0.1 8.9 ± 0.1 8.9 ± 0.0
acteristic of the proposed instrument concerns to the use of a
phototransistor with spectral sensitivity in the visible region as
R.S.D. 0.8% 0.5% 0.7%
phototransductor.
G. Veras et al. / Talanta 77 (2009) 1155–1159 1159

The MLCS was successfully applied to the determination of food References


colorants (sunset yellow, tartrazine, allura red and brilliant blue)
in synthetic samples and Fe2+ in restorative oral solutions. In all [1] H. Flaschka, C. Mckeithan, R. Barnes, Anal. Lett. 6 (1973) 585.
[2] P.K. Dasgupta, I.Y. Eom, K.J. Morris, J. Li, Anal. Chim. Acta 500 (2003) 337.
applications, the proposed instrument lead to results similar to the [3] E.N. Gaião, E.P. Medeiros, W.S. Lyra, P.N.T. Moreira, P.C. Vasconcelos, E.C. Silva,
obtained by the commercial instruments employed for compari- M.C.U. Araújo, Quim. Nova 28 (2005) 1102.
son. Therefore, the proposed instrument offers an economically [4] Y. Shimazaki, S. Watanabe, M. Takahashi, M. Iwatsuki, Anal. Sci. 16 (2000)
1091.
viable alternative for spectrophotometric chemical analysis in small [5] Q. Li, K.J. Morris, P.K. Dasgupta, I.M. Raimundo Jr., H. Temkin, Anal. Chim. Acta
routine, research and/or teaching laboratories. 479 (2003) 151.
[6] M.A. Cavalcante, A. Benedetto, Rev. Bras. Ensino Física 21 (1999) 437.
[7] IUPAC—Nomenclature, symbols, units and their usage in spectrochemical anal-
Acknowledgement ysis. II. Data interpretation. Pure Appl. Chem. 45 (1976) 99.
[8] P. Bruce, P. Minkkinen, M.-L. Riekkola, Microchip. Acta 128 (1998) 93.
The authors thank the Brazilian agency CNPq for scholarship.

You might also like