Temperature-stable parallel-plate dielectric cell for broadband liquid impedance measurements

Brian A. Mazzeo, Satyan Chandra, Brett L. Mellor, and Jesus Arellano Citation: Rev. Sci. Instrum. 81, 125103 (2010); doi: 10.1063/1.3509388 View online: http://dx.doi.org/10.1063/1.3509388 View Table of Contents: http://rsi.aip.org/resource/1/RSINAK/v81/i12 Published by the American Institute of Physics.

Related Articles
Admittance of Au/1,4-benzenedithiol/Au single-molecule junctions Appl. Phys. Lett. 101, 253510 (2012) Measurement of the electrical resistivity and Hall coefficient at high temperatures Rev. Sci. Instrum. 83, 123902 (2012) Scanning nonlinear dielectric microscopy observation of accumulated charges in metal-SiO2-SiN-SiO2-Si flash memory by detecting higher-order nonlinear permittivity Appl. Phys. Lett. 101, 242101 (2012) BX90: A new diamond anvil cell design for X-ray diffraction and optical measurements Rev. Sci. Instrum. 83, 125102 (2012) Experimental verification of the 3-step model of photoemission for energy spread and emittance measurements of copper and CsBr-coated copper photocathodes suitable for free electron laser applications Appl. Phys. Lett. 101, 231103 (2012)

Additional information on Rev. Sci. Instrum.

Journal Homepage: http://rsi.aip.org Journal Information: http://rsi.aip.org/about/about_the_journal Top downloads: http://rsi.aip.org/features/most_downloaded Information for Authors: http://rsi.aip.org/authors

Downloaded 20 Dec 2012 to 124.124.48.3. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions

REVIEW OF SCIENTIFIC INSTRUMENTS 81, 125103 (2010)

Temperature-stable parallel-plate dielectric cell for broadband liquid impedance measurements

Brian A. Mazzeo,a) Satyan Chandra, Brett L. Mellor, and Jesus Arellano
Department of Electrical and Computer Engineering, Brigham Young University, Clyde Building, Provo, Utah 84602, USA

(Received 1 July 2010; accepted 10 October 2010; published online 17 December 2010) A liquid impedance cell for broadband impedance measurements up to 110 MHz is presented. The design incorporates temperature control and minimizes parasitic capacitance and inductance. The cell is simple to fabricate and uses chemically resistant materials, stainless steel, and Teon. This dielectric cell can be used in a variety of liquid measurements, particularly those related to impedance measurements of biological objects in solution. Temperature control is illustrated in measurements of the permittivity of deionized water from 5 to 55 o C. Numerical tting procedures employed on the relaxation curves indicate good agreement with previous studies on beta-lactoglobulin and hen lysozyme. Titration capability is demonstrated through dielectric titration of hen lysozyme and betalactoglobulin. 2010 American Institute of Physics. [doi:10.1063/1.3509388]
I. INTRODUCTION

Liquid electrical measurements can reveal a broad variety of interesting and important phenomena from physical and chemical processes occurring within solutions.1 Dielectric spectroscopy has historically been used for protein measurements to compute the dipole moment in varying solution conditions.2 Because of this, new developments in electrical equipment and measurement cells over the past century have greatly enhanced the ability of the researcher to access important dielectric properties.3, 4 Computer-controlled impedance and network analyzers have increased measurement resolution to distinguish smaller changes in dielectric properties as well as sampling speeds. To push the upper frequency limits in the GHz range, waveguide and coaxial techniques are often employed; designs have included open and closed cells.5, 6 On the low-frequency side, the MHz range and below, electrode polarization is the dominant parasitic inuence. Electrode polarization is caused by the formation of a layer of charge near the surface of the metal electrode.7, 8 Because it is in series with the measurement of the solution properties, it frequently disrupts accurate measurements of the permittivity of proteins. It has historically been a major obstacle to accurate measurements and continues to complicate measurement interpretation.9 For many years, the only solution to electrode polarization was seen as four-electrode dielectric spectrometers.10 Many four-electrode arrangements were built on this assumption (see Ref. 11 and references therein). Recently, the supposed advantage of this electrode arrangements has been questioned.12, 13 These studies have shown that electrode polarization is a nuisance for many different arrangements and is not easily mitigated. The only consistent way to reduce the effects of this series polarization is to increase the electrode spacing between the parallel plates.13 Even then, numerical techniques must be used
a) Electronic mail: bmazzeo@ee.byu.edu.

to postprocess the data to further reduce the polarization contribution.14 Research on low-frequency cells and electrode polarization is active. Examples of low-frequency cells include variable electrode separation15 and blocking electrodes.16 Correspondingly, electrode polarization problems continually receive attention because of their signicant inuence on measurements.7, 12, 17, 18 Reducing the inuence of polarization enables better understanding of fundamental electrical properties of biological systems. Platinum black electrodes are frequently used to minimize the electrode impedance, but because of regeneration requirements and corresponding variability, research continues into different electrode materials.19 From the scientic literature, it appears that no one dielectric cell has perfect properties across the entire obtainable spectra; in many cases, multiple cells and instruments must be used to perform the measurements in different ranges. This motivates dielectric cells constructed for particular targets. This paper addresses a new dielectric cell that effectively spans the critical MHz range where the dielectric relaxations due to overall protein dipole moments are visible. Electrical measurements of protein interaction processes in solution have been examined recently.20 The primary, or beta, relaxation of a protein molecule typically occurs in the MHz range due to the hydrodynamic properties of the molecule in solution. In this frequency range, the major obstacles to accurate measurements of proteins in solution are conductivity of the solution, electrode polarization, temperature variation, broadband characteristics, and titration capability. The admittance of a parallel-plate capacitor with a solution of conductivity, , and relative permittivity, , can be given by Y = A ( + i 0 ) , d (1)

where A is the area of the electrode, d is the distance between the plates, is the applied frequency of measurement, and 0 is the permittivity of free space.11 To resolve the
2010 American Institute of Physics

0034-6748/2010/81(12)/125103/5/$30.00

81, 125103-1

Downloaded 20 Dec 2012 to 124.124.48.3. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions

125103-2

Mazzeo et al.

Rev. Sci. Instrum. 81, 125103 (2010)

permittivity of the solution requires accurate determination of the phase between the real and imaginary parts of the impedance. High conductivity solutions, thus, reduce the phase angle and make it difcult to measure the permittivity. This is one reason that most dielectric spectroscopy experiments are performed in solutions with an ionic strength less than 1 mM. At higher frequencies, the phase angle is larger due to the frequency-dependent impedance of the capacitance and it is easier to resolve permittivity. The low-frequency impedance data is also corrupted by contributions from the electrodes. In precision protein experiments, the dielectric increment associated with proteins is an order of magnitude smaller than the background permittivity of the solution. In this case, stability of the medium is critical so the small permittivity signal can be extracted. The permittivity of water is noticeably dependent on temperature.21 Additionally, the relaxation characteristics of proteins are also dependent on temperature.22 Because of this dependence, strict controls are necessary to keep the solution stable. Many commercially available cells do this by complete immersion into a bath. However, this may prevent the introduction of reaction agents into the cell without disturbing the nely tuned characteristics of the cell. Alternative methods of temperature stabilization, such as ow-through electrodes, may be employed.23 To perform dielectric spectroscopy experiments using a single instrument, it is convenient to perform measurements across a broad range of frequencies. A precision impedance analyzer, the Agilent 4294A, has a range of 40 Hz110 MHz. A broad range of phenomena are accessible, particularly protein beta relaxation. However, at high frequencies the inductance of the cell combines with the capacitance to form a resonant mode at the high frequencies. Cells in the past have overcome this by reducing the length of the leads to the cell and optimizing the plate geometry.23, 24 Titration capability is important for dielectric measurements because it allows the baseline to be established for the electrode polarization and solvent permittivity. Removing and adding liquids then are perturbations that can be measured relative to the baseline. In these measurements, an assumption is made that the electrode polarization is relatively constant during the experiment. However, this differential method can resolve lower concentrations of protein by suppressing background parasitic contributions. Computer control is necessary to take repeated measurements and to plot time-resolved studies of protein interaction. Dielectric cells are constructed of materials that can be autoclaved and cleaned with solvents to remove organic contaminants. Teon is a popular choice for these applications due to thermal stability and chemical resistance. The cell in this paper was designed with these considerations in mind to provide a stable platform for dielectric titration of proteins. It is a signicant improvement over past cells used for this purpose.11, 20
II. DIELECTRIC CELL DESIGN

the entire upper range of the impedance analyzer, up to 110 MHz. The base material for the analyzer was Teon, which has low thermal conductivity 0.26 W/(K m).25 The electrodes themselves were milled from 3/8 in. diameter stainless steel, ASTM Grade 304. A slight chamfer at the electrode edge was added to ease insertion and tight sealing against the Teon. The surfaces were ground with 600, 800, and 1200 grit silicon carbide abrasive discs in a Spectrum System 2000 (LECO Corp.). Deionized (DI) water owed from a NESLAB RTE-40 thermal bath to provide temperature stability. The design takes advantage of the high thermal conductivity and electrical conductivity of the stainless steel for stabilization and sensing. A rubber bung is used to seal the inner cavity. A small air gap between the liquid and the bung is sufcient to accommodate liquid expansion and contraction during thermal cycling. Screws through tapped holes of size 256 were fastened through the Teon to provide electrical connection to the inner stainless-steel electrodes. The inner, parallel-plate cavity of the dielectric cell holds 800900 l of uid. The reduction of metal usage to only parallel electrode surfaces is notable in this design to reduce the parasitic capacitance of this cell. The relatively large spacing between the electrodes also reduces electrode polarization in series with the solution. The schematic for this cell is shown in Fig. 1. The entire cell was mounted on a plastic plate connected to the four terminals of the 4294A impedance analyzer. The screws had washers and bolts that held wire that was soldered to crossbars between the low current and potential terminals and the high current and potential terminals. In this fashion, the effective cable length from the terminals of the analyzer to the cell was minimized. Figure 2 shows a photograph of the completed cell mounted to the Agilent 4294A. A LABVIEW computer interface controlled the impedance analyzer through a GPIB connection. The RTE-40 thermal bath was also controlled via the LABVIEW computer interface through a serial connection. In this fashion, the solution temperature could be accurately controlled over the entire range of the bath and impedance measurements could be taken at selected temperatures. Generally, sweeps were performed over 401 logarithmically spaced points from 40 Hz to 110 MHz with settings of a bandwidth of 5 (highest resolution) and oscillator strength of 500 mV.

III. EXPERIMENTS AND DISCUSSION

Given the above requirements, a new cell was constructed that allows the measurement of protein characteristics over

To verify the accurate temperature control and ability to resolve protein relaxations up to 110 MHz, multiple experiments were performed on standard uids and proteins. These experiments demonstrated that the spectrometer could indeed measure temperature-dependent solution permittivity and protein titrations. To establish a baseline cell constant, the following approach was used. The water bath was set to 25 o C and was allowed to stabilize for 15 min. The capacitance at 1 MHz was measured with the cell empty and then with 800 l of DI water added followed by 15 min of stabilization time. Using these values of capacitance, C, the cell constant, , and the

Downloaded 20 Dec 2012 to 124.124.48.3. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions

125103-3

Mazzeo et al.
STAINLESS STEEL GRADE 304 ELECTRODES DIAMETER 9.5 X LENGTH 12.7

Rev. Sci. Instrum. 81, 125103 (2010)

FRONT

17.8

SIDE

35.0

11.2

0 15.6

18.0

SCREWS ON REAR THREAD 2-56

17.8

TOP
9.2

8.0

27.0

76.0

FIG. 1. Schematic of dielectric cell designed for protein titration. Dimensions are in milimeter. DI water ows through ports in the front and sides. This ow by the interior electrodes maintains constant temperature in the liquid cavity.

parasitic capacitance, CP , were determined using values of 1 and 78.368 for the DI water through the formula, C = + C P . (2)

The measured cell constant was found to be 0.0494 pF and the parasitic capacitance was calculated to be 0.712 pF at 1 MHz. To test the ability of the apparatus to scale through a range of temperatures, the temperature was scaled from 5 to 55 o C in 5o increments. The bath was held at the set temperature for 15 min before multiple frequency sweeps were performed. The permittivity of the water at 1 MHz as measured and compared to standards is shown in Fig. 3. Indeed, the agreement is very good, particularly within the 15o 40o range, which covers most physiological temperatures. Depar-

ture from the ideal measurements at the temperature extremes is due to the stronger thermal gradients from the interior to the exterior of the cell. At low temperatures, this is most noticeable. Next, a titration of beta-lactoglobulin and hen lysozyme was performed in the dielectric cell. Beta-lactoglobulin (L3908) and hen lysozyme (L6876) were obtained from Sigma and reconstituted in 0.1 mM HCl at a concentration of 20 mg/ml. The solutions were mixed and stored in microcentrifuge tubes. The cell was rinsed with ethanol and DI water

90 Measured Ideal 85

Permittivity

80

75

70

65

10

20 30 40 Temperature [C]

50

60

FIG. 2. (Color online) Photograph of the dielectric cell connected to the 4294A Precision Impedance Analyzer. The tubes carrying DI water to and from the NESLAB RTE-40 circulating bath are visible.

FIG. 3. (Color online) Permittivity versus temperature from 5 to 55 o C. The plot of the ideal temperature is taken from (Ref. 21). The dielectric cell has stable temperature readings through this range, with small departure at the temperature extremes.

Downloaded 20 Dec 2012 to 124.124.48.3. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions

125103-4

Mazzeo et al.

Rev. Sci. Instrum. 81, 125103 (2010)

Differential, Normalized Permittivity

and allowed to dry. The 0.1 mM HCl solution was placed in the cell and baseline measurements were taken at 25 o C. The sweeps were postprocessed in MATLAB and analyzed. A shift in the capacitance due to the impedance analyzer range shift was removed. Additionally, the data was normalized to 110 MHz. Differential measurements were then taken with respect to the original 0.1 mM HCl baseline to remove effects of electrode polarization. Least-squares ttings were done with the function lsqcurvet, with frequencies expressed logarithmically. Measurements of permittivity were tted to the real part of a single relaxation ColeCole curve: = + , 1 + j ( )(1) (3)

HENL additions

BLG additions 30 20 10 0 5 6 Frequency [log (Hz)]

10

where is the high frequency permittivity, is the change in permittivity, is the Cole parameter describing the spread of the relaxation,26 is the relaxation time, and j = 1. The electric dipole moment is related to through the Oncley formula: = 2 Mkb T 0 , Ng (4)

7 8 25

50

75 Time [Min]

100

125

FIG. 4. (Color online) Plot of dielectric titration of beta-lactoglobulin and hen lysozyme over time at 25 o C. The relaxations of the proteins are visible in this plot. It is noticeable that the time constant of the combined proteins shifts considerably on addition of the hen lysozyme, corresponding to increased hydrodynamic volume of the complex.

where is the dipole moment, M is the protein molecular weight in kilodaltons, kb is the Boltzmann constant, T is the temperature in Kelvin, 0 is permittivity of free space, N is Avogadros number, g is the correlation parameter assumed to be 1 for dilute protein solutions,22 and = limc0 /c is the dielectric increment where c is the protein concentration in mg/ml. Assuming the protein is roughly spherical, the effective hydrodynamic radius of the protein can be estimated by the formula, = 4 a 3 , kb T (5)

where a is the effective hydrodynamic radius and is the viscosity of the solvent.27 Using Eq. (5), the hydrodynamic radii of BLG and HENL were estimated to be 26 and 20 , respectively, these values being similar to structural data deposited in the Protein Data Bank.28, 29 Estimated dipole moments of BLG and HENL were 800 40 D and 270 20 D. The value for BLG is in good agreement with the value measured by the pioneering work of Ferry and Oncley of 720 D.30 The measurements of Bonincontro et al.31 on lyszozyme gave dipole moments around 400 D in water, around 300 D in water from pH 4 to pH 6,32 and 210 D in water in other experiments,33 indicating our results are within the expected range. Additional theoretical and experimental works, like in this study, will help to establish better dipole moment standards for proteins. Continuous dielectric measurements were then taken of a protein titration. A pipettor was used to remove 60 l of liquid from the dielectric cell. Here, 60 l of the concentrated BLG solution was then added to the cell to form a concentration of 1.5 mg/ml. After each titration, the solution was allowed to stabilize for 20 min to reach thermal and chemical equilibrium. Here, 60 l of liquid was again removed and 60 l of the concentrated BLG solution was added to form 3 mg/ml BLG solution. Then 60 l of the solution was removed and

60 l of concentrated HENL solution was added. This step of removing 60 l of solution and adding 60 l of concentrated HENL was repeated twice. The results of this titration are shown in Fig. 4 and clearly indicate the interaction between these two proteins.34 When just one protein is present, the dielectric relaxation for that protein is strongly visible. When the complementary protein is added, the dielectric relaxation shifts to lower frequencies. This shift takes place because the aggregate is now much bigger than the individual proteins that constitute the aggregate. The increased hydrodynamic volume impedes the rotation of the aggregate.4 It is also evident from the data that electrode polarization is affecting the measurements as the concentration of dissolved ions in the solution increases. However, this cell has much lower polarization below 1 MHz as compared with other cells.20, 32 Further experiments varying the solution and protein conditions will elucidate how the binding reaction takes place in the solution. These experiments clearly indicate that this dielectric cell can successfully stabilize temperature for dielectric spectroscopy experiments and can resolve protein titrations in solution. Future experiments will elucidate the protein association mechanisms within the solution. This should allow for accurate and repeatable measurements of protein solutions for comparison with theoretical determination of protein electrical parameters.

IV. CONCLUSION

A new dielectric cell apparatus is presented and used for accurate titration of protein solutions. It is an easy-to-use system that takes advantage of the thermal and electrical properties of stainless steel to provide reduced electrode polarization in permittivity data at low frequencies and measurements of protein relaxations up to 110 MHz.

Downloaded 20 Dec 2012 to 124.124.48.3. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions

125103-5

Mazzeo et al.
16 F.

Rev. Sci. Instrum. 81, 125103 (2010) Pizzitutti and F. Bruni, Rev. Sci. Instrum. 72, 2502 (2001). D. Hollingsworth and D. A. Saville, J. Colloid Interface Sci. 257, 65 (2003). 18 M. Tirado and C. Grosse, Colloids Surf. A 222, 293 (2003). 19 P. Mirtaheri, S. Grimnes, and G. Martinsen, IEEE Trans. Biomed. Eng. 52, 2093 (2005). 20 B. A. Mazzeo and A. J. Flewitt, Appl. Phys. Lett. 90, 123901 (2007). 21 W. J. Ellison, K. Lamkaouchi, and J. M. Moreau, J. Mol. Liq. 68, 171 (1996). 22 R. Pethig, Dielectric and Electronic Properties of Biological Materials (Wiley, New York, 1979). 23 C. G. Essex, G. P. South, R. J. Sheppard, and E. H. Grant, J. Phys. E 8, 385 (1975). 24 B. P. Jordan and E. H. Grant, J. Phys. E 3, 764 (1970). 25 CRC Handbook of Chemistry and Physics, 91st ed. (Internet Version 2011), edited by W. M. Haynes (CRC Press/Taylor and Francis, Boca Raton, FL). 26 K. S. Cole and R. H. Cole, J. Chem. Phys. 9, 341 (1941). 27 P. Debye, Polar Molecules (Chemical Catalog, New York, 1929). 28 S. Brownlow, J. H. M. Cabral, R. Cooper, D. R. Flower, S. J. Yewdall, I. Polikarpov, A. C. T. North, and L. Sawyer, Structure 5, 481 (1997). 29 H. Schwalbe, S. B. Grimshaw, A. Spencer, M. Buck, J. Boyd, C. M. Dobson, C. Redeld, and L. J. Smith, Protein Sci. 10, 677 (2001). 30 J. D. Ferry and J. L. Oncley, J. Am. Chem. Soc. 63, 272 (1941). 31 A. Bonincontro, A. De Francesco, M. Matzeu, G. Onori, and A. Santucci, Colloids Surf., B 10, 105 (1997). 32 A. Bonincontro, A. De Francesco, and G. Onori, Colloids Surf., B 12, 1 (1998). 33 A. Bonincontro, S. Cinelli, G. Onori, and A. Stravato, Biophys. J. 86, 1118 (2004). 34 M. Nigen, T. Croguennec, D. Renard, and S. Bouhallab, Biochem. 46, 1248 (2007).
17 A.

ACKNOWLEDGMENTS

The authors gratefully acknowledge the support of the College of Engineering and the Ofce of Research and Creative Activities at Brigham Young University for research initiation funds to support this work. The authors thank Don Dawson at Brigham Young University for fabrication assistance and thank Dr. Tracy Nelson at Brigham Young University for providing polishing materials for the electrodes.
Kaatze and Y. Feldman, Meas. Sci. Techol. 17, R17 (2006). L. Oncley, Biophys. Chem. 100, 151 (2003). 3 J. Wyman Jr., J. Bio. Chem. 90, 443 (1931). 4 A. Oleinikova, P. Sasisanker, and H. Weingartner, J. Phys. Chem. B 108, 8467 (2004). 5 O. Gttmann, U. Kaatze, and P. Petong, Meas. Sci. Techol. 7, 525 (1996). 6 K. Folgero, IEEE Trans. Instrum. Meas. 47, 881 (1998). 7 F. Bordi, C. Cametti, and T. Gili, Bioelectrochem. 54, 53 (2001). 8 A. D. Hollingsworth and D. A. Saville, J. Colloid Interface Sci. 280, 279 (2004). 9 W. Scheider, J. Phys. Chem. 79, 127 (1975). 10 H. P. Schwan and C. D. Ferris, Rev. Sci. Instrum. 39, 481 (1968). 11 B. A. Mazzeo and A. J. Flewitt, J. Appl. Phys. 102, 104106 (2007). 12 S. Grimnes and O. G. Martinsen, J. Phys. D 40, 9 (2007). 13 B. A. Mazzeo, J. Appl. Phys. 105, 094106 (2009). 14 Y. Feldman, E. Polygalov, I. Ermolina, Y. Polevaya, and B. Tsentsiper, Meas. Sci. Techol. 12, 1355 (2001). 15 M. C. Tirado, F. J. Arroyo, A. V. Delgado, and C. Grosse, J. Colloid Interface Sci. 227, 141 (2000).
2 J. 1 U.

Downloaded 20 Dec 2012 to 124.124.48.3. Redistribution subject to AIP license or copyright; see http://rsi.aip.org/about/rights_and_permissions