Direct Alkalinity Detection with Ion-Selective Chronopotentiometry

Subscriber access provided by Penn State | University Libraries

Article

Direct Alkalinity Detection with Ion-Selective Chronopotentiometry Majid Ghahraman Afshar, Gastón A. Crespo, Xiaojiang Xie, and Eric Bakker Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/ac500968c • Publication Date (Web): 27 May 2014 Downloaded from http://pubs.acs.org on June 5, 2014

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Direct Alkalinity Detection with Ion-Selective Chronopotentiometry Majid Ghahraman Afshar, Gastón A. Crespo, Xioajiang Xie and Eric Bakker* Department of Inorganic and Analytical Chemistry, University of Geneva, Quai Ernest-Ansermet 30, CH1211 Geneva, Switzerland.

KEYWORDS: Alkalinity, Perm-selective Membranes, Flash Chronopotentiometry, IonDepletion

ACS Paragon Plus Environment

1


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 32

ABSTRACT We explore the possibility to directly measure pH and alkalinity in the sample with the same sensor by imposing an outward flux of hydrogen ions from an ion-selective membrane to the sample solution by an applied current. The membrane consists of a polypropylene supported liquid membrane doped with a hydrogen ionophore (chromoionophore I), ion exchanger (KTFBP) and lipophilic electrolyte (ETH 500). While the sample pH is measured at zero current, alkalinity is assessed by chronopotentiometry at anodic current. Hydrogen ions expelled from the membrane undergo acid–base solution chemistry and protonate available base in the diffusion layer. With time, base species start to be depleted owing to the constant imposed hydrogen ion flux from the membrane, and a local pH change occurs at a transition time. This pH change (potential readout) is correlated to the concentration of the base in solution. As in traditional chronopotentiometry, the observed square root of transition time (τ) was found to be linear in the concentration range of 0.1 mM to 1 mM, using the bases trisaminomethane, ammonia, carbonate, hydroxide, hydrogen phosphate and borate. Numerical simulations were used to predict the concentration profiles and the chronopotentiograms, allowing on discussing the possible limitations of the proposed method and its comparison with volumetric titrations of alkalinity. Finally, P-alkalinity level is measured in a river sample to demonstrate the analytical usefulness of the proposed method. As a result of these preliminary results, we believe that this approach may become useful for the in-situ determination of P-alkalinity in a range of matrices.

.


2

Page 3 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION Alkalinity and pH are routinely measured as estimators of water and food quality.1-2 Typically, these determinations are performed in laboratories using traditional analytical methodologies. On one hand, pH is a measure of the concentration of hydrogen ions in solution and it is usually determined by potentiometry using ion-selective membranes.3-5 On the other hand, alkalinity describes the capacity of the sample to neutralize acids.6-7 Determination of total alkalinity is accomplished with titrimetric analysis, where the equivalent of acid used to reach a pH of 4.2 correspond to the total alkalinity of the sample.8 In environmental samples (i.e., tap water, river water, etc.) the predominant species are bicarbonate and carbonate that determine the pH of the sample. For this reason, two main alkalinity parameters are used in order to describe the sample. P-alkalinity refers to carbonate concentration (acid titration with phenolphthalein as visual indicator) whereas M-alkalinity refers to the sum of carbonate and bicarbonate concentration (acid titration with methyl-orange as visual indicator).8 In view of in-situ determinations, the titrant is produced by electrochemical methods, typically from a precursor compound as a result of a redox process at the working electrode.9 Generally, water is oxidized at the platinum electrode to generate hydrogen ions. The amount of titrant is carefully adjusted either by the applied current or voltage. In fact, coulometry and amperometry are the most used methodologies to perform such measurements.9-16 Evidently, this method is based on a bulk sample titration and convective mass transport is required during the experiment. In a different approach, ion selective membranes were also used to perform bulk titrations. The first reports were introduced by Horvai et al. for determining hydrochloric acid and sodium hydroxide in a limited concentration range.17-18 The experimental setup involved a wall-jet cell to reduce the diffusion time. In subsequent work, the selective coulometric release of calcium from


3


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 32

ion selective polymer membranes was explored. While such bulk sample titrations promise to be very accurate, they also necessitate the use of thin layer sampling compartments, fluidic systems or require precise aliquots and appear at this stage less convenient for continuous measurements in unmodified samples in applications such as environmental monitoring.19-20 More

recently,

ionophore-based

membranes

have

been

interrogated

by

flash-

chronopotentiometry in our group.21 In this methodology, anodic or cathodic curents are applied to the membrane to locally deplete anions or cations at the sample–membrane interface. A flashchronopotentiometric method was successful proposed for the detection of calcium, potassium, proton and protamine.22-23 In preliminary work, an anodic pulse applied to a permselective hydrogen ion-selective membrane was found to result in the controlled release of protons from the membrane to the aqueous diffusion layer.24 The square root of transition time gave a linear dependence on the consumed concentration of base, trisaminomethane (Tris), in the sample. In view of extending the previous concept, a permselective membrane is used here in chronopotentiometric and potentiometric tandem sensing mode for total alkalinity and pH determinations. While two separate sensors could be used for pH and alkalinity, an integration of both allows one to evaluate the proper functioning of the membrane after applying current and simplify the final device for easier miniaturization. In chronopotentiometric mode, a flux of protons is established from the membrane into the sample. This process locally titrates the base in the sample by the released proton. Total alkalinity (or P-alkalinity) and the diffusion coefficients of a number of bases (borate, phosphate, ammonia, carbonate, hydroxide and tris) are determined while the same permselective membrane is used as a pH electrode in zero current potentiometry. Other types of membrane such as Nafion or glass are not suitable for this application, since a sufficiently high outgoing proton flux cannot be established because of a lack


4

Page 5 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of selectivity with the former and lack of conductivity with the latter material. Ionophore-based membranes are able to generate proton fluxes with sufficient selectivity and amplitude and are explored in this work.


5


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 32

EXPERIMENTAL SECTION Reagents and solutions Potassium

tetrakis[3,5-bis(trifluoromethyl)phenyl]borate

chlorophenyl)borate

tetradodecylammonium

salt

(ETH

500),

(KTFPB),

tetrakis(4-

chromoionophore

I,

2-

nitrophenyloctylether (o-NPOE), tris(hydroxymethyl)aminomethane (Tris), acetic acid, sodium acetate, sodium chloride, sodium hydroxide (1 M) and tetrahydrofuran (THF) were purchased from Sigma-Aldrich (analytical grade). Electrochemical equipment A double-junction Ag/AgCl/3M KCl/1M LiOAc reference electrode was used in potentiometric and chronopotentiometric measurements (Mettler-Toledo AG, Schwerzenbach, Switzerland). Electrode bodies (Oesch Sensor Technology, Sargans, Switzerland) were used to mount the polymeric membranes. A platinum working rod (3.2 cm2 surface area) served as a counter electrode. Selectivity coefficients were determined by zero current potentiometry employing high impedance input 16-channel EMF monitor (Lawson Laboratories, Inc., Malvern, PA). Potentiometric and chronopotentiometric measurements were performed with an Autolab PGSTAT302N (MULTI 16 module, Metrohm Autolab, Utrecht, The Netherlands) that allows one to read out up to 16 working electrodes placed in the same electrochemical cell. A Faraday cage was used to protect the system from undesired noise.

Membrane preparation


6

Page 7 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Porous polypropylene (PP) membranes (Celgard, 0.237 cm2 surface area, 25 µm thickness, and kindly provided by Membrane Wuppertal, Germany) were used as supporting material. The membranes were washed with THF for 10 min to remove any possible contaminants. When the membrane was found to be completely dry (in a matter of seconds), an excess volume of 3 µL of the cocktail solution was deposited on it (see cocktail preparation below). The impregnation of the cocktail was found to be instantaneous; however, the membrane was let in the Petri Dish for ca. 10 min to ensure a homogenous and reproducible impregnation of the pores. The pore filling solution composition is assumed to remain identical to the initial THF-free cocktail. Afterwards, the membrane was conditioned in the buffer solution for 40 min. Finally, the membrane was mounted in the electrode body. The inner compartment was filled with 10 mM acetic acid/10 mM sodium acetate buffer in 10 mM of NaCl for all experiments. All solutions were prepared in 10 mM of NaCl as background electrolyte to avoid migration process (higher current densities might require higher background electrolyte concentration). The cocktail H1 for alkalinity measurements contained 120 mmol kg-1 of Ionophore I (mmol per kg of membrane without THF), 60 mmol kg-1 of KTFPB, 90 mmol kg-1 of ETH 500, 190 mg of o-NPOE and 1 mL of THF. Note that THF was only used to enhance the solubility of the solid compounds into the plasticizer and was removed by evaporation before casting the membranes. Electrochemical protocol A fully automatized (NOVA software, Autolab) method consists of three steps: i) Open circuit potential determination for 5-s at a sampling rate of 10 Hz, (no current flow through the electrochemical cell), ii) Anodic constant current pulse for 5 s at potential sampling rate of 500 Hz, iii) Constant potential pulse for 60-s at sampling rate of 500 Hz (same potential as recorded in i), regeneration membrane step). The time interval between each step was 1 ms. The duration


7


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 32

of the anodic pulse (5s) was chosen after experimental optimization. For these determinations, a current pulse of 20 µA was normally used. Sample Solutions Stock solutions were prepared from a 1M solution of each base, adjusting their pH by adding 1 M HCl solution. The pH of the stock solutions were: 7.9 (hydrogen phosphate), 9.4 (ammonia), 9.5 (tris) 9.5 and 8.6 (borate) and 8.7 (carbonate). All the pH values were measured with a Metrohm pH electrode and pH meter. Calibration curves were obtained by adding precise volumes of stock solution to 100 mL of milliQ-purified water.


8

Page 9 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


RESULTS AND DISCUSSION The objective of this work is to characterize an electrochemical sensing tool to directly determine alkalinity and pH. For this purpose, we selected a permselective membrane that can be sequentially operated by chronopotentiometry and zero current potentiometry. Figure 1 schematically illustrates the operational principle of a proton ion selective membrane sandwiched between an inner solution and a sample solution. The membrane contains chromoionophore I (L), cation exchanger (R−), lipophilic electrolyte (R+R−) and plasticizer (oNPOE). The optimal membrane composition contains a 100% molar excess of chromoionophore I over the cation exchanger. Typically, the presence of cation exchanger suppresses the extraction of counterions from the backside of the membrane.25 In addition, the membrane contains a high concentration of ionophore (120 mmol kg-1) to increase the proton release capacity. Different ratios of ionophore / cation exchanger were explored with 120:60 mmol kg-1 most suitable for alkalinity determination (data not shown). In the proposed method, hydrogen ions are selectively transported across the membrane. Therefore, a selective incorporation of protons is required at the inner interface in order to replace the released proton at the sample interface. Because the selectivity coefficients for membranes containing chromoionophore I are relatively high (values log KpotH,Na = -10.9) 26 , the incorporation of ions other than hydrogen ions must be insignificant in the working time scale. Additionally, we aim to utilize membranes that also work in potentiometry and adequate selectivity is required for measuring the pH in different background electrolytes solutions. The use of other types of membrane such as Nafion will not exhibit selective transport behavior for hydrogen ions as evidenced by potentiometric characterization. 27


9


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 32

The difference of potentials (ΔE) between the working electrode (WE) and the commercial reference electrode (RE) is related to the proton activity changes (pH) in the course of the potentiometric experiment (zero current). When a galvanostatic pulse (an anodic current) of 5-s duration is applied to the electrochemical cell, hydrogen ions are transported across the membrane from the inner solution to the sample. The established proton flux (JH) at the samplemembrane interface (position x,y=0 respectively) is well described by the equation 1 (x and y correspond to the coordinates in the aqueous and organic phase respectively) where i is the applied current, A is the area of the membrane, DHaq, DHLm, cHaq and cHLm are the diffusion coefficients and concentrations of proton and proton-ionophore adduct in their respective phases, labeled as aq and m.

JH =

⎡ ∂c aq ⎤ ⎡ ∂c m ⎤ i = − DHaq ⎢ H ⎥ = DHLm ⎢ HL ⎥ FA ⎣ ∂ y ⎦ y=0 ⎣ ∂x ⎦ x=0

(1)

If a base (proton acceptor) is present in the sample solution, it can be locally titrated at the sample-membrane interface. Assuming that the protonation reaction is sufficiently fast and diffusion is along one dimension only, the concentration profile of the base can be described in terms of the applied current as follows (JB is flux of the base and DBaq corresponds to the diffusion coefficient of B in the aqueous phase):

JB =

⎡ ∂c aq ⎤ i = DBaq ⎢ B ⎥ FA ⎣ ∂x ⎦ x=0

(2)

A relationship between the concentrations of hydrogen ions (cH) and protonated (cBH) and nonprotonated (cB) base at the membrane surface of the sample is found using the mass balance (with ci is the initial concentration of the base) and the acidity constant (Ka, strictly activities instead of concentrations) as follows:


10

Page 11 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


cB + cBH = ci

Ka =

(3)

cB cH +

(4)

cBH +

At the transition time of t = τ the concentration of non-protonated base approaches zero. At this time (τ) one may formulate the Sand equation after Laplace transformation of Eqn 2 as:

JB =

i 1 DBπ * = cB FA 2 τ

(5)

This equation is valid for cB (x=0) = 0 whereas cB* correspond to the bulk concentration of the base. Time dependent phenomena are simulated by the finite difference method. For this purpose we use the continuity equation as:

∂C B (x) ∂J B (x) = ∂t ∂x

(6)

In a numerical simulation with equal distance steps (d), we assume linear gradients from any step ν to step ν+1.28 Equation 2 is rewritten as:

J B ,v / v +1 (t ) = DB

cB (v, t ) − ca (v + 1, t ) d

(7)

where JB,v/v+1(t) is the flux from v-th to the (v+1)-th element and cB(v,t) is the concentration in the respective element. The continuity equation (6) is rewritten as:

cB ,v (t + Δt ) − cB ,v (t ) Δt

=

J B ,v −1/ v (t ) − J B ,v / v +1 (t ) d

(8)

Inserting Eqn. (7) into Eqn. (8) for the indicated positions gives: cB ,v (t + Δt ) − cB ,v (t ) Δt

⎧ c (v − 1, t ) − cB (v, t ) ⎫ ⎧ cB (v, t ) − cB (v + 1, t ) ⎫ = DB ⎨ B ⎬ − DB ⎨ ⎬ 2 d d2 ⎩ ⎭ ⎩ ⎭

(9)

Which is solved for each position v for the concentration of the following time step:


11


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

cB ,v (t + Δt ) = cB ,v (t ) +

Δt DB {cB (v − 1, t ) − 2cB (v, t ) + cB (v + 1, t )} d2

Page 12 of 32

(10)

Similarly, Eqn. 10 can be written for BH+ and H+ species before the transition time using the mass balance and the acidity constant (Eqn. 3 and 4). After the transition time, hydrogen ions are still released from the membrane as hydronium ions, and the concentration profile is described as follows:

JH =

⎡ ∂c ⎤ i = − DH ⎢ H ⎥ FA ⎣ ∂x ⎦ x=0

cH ,v (t + Δt ) = cH ,v (t ) +

(11)

Δt DH {cH (v − 1, t ) − 2cH (v, t ) + cH (v + 1, t )} d2

(12)

As a result, when a current is applied to the membrane, the outgoing ion flux is exclusively related to hydrogen ions. With membranes of sufficient selectivity, the uncertainty in the proton flux amplitude is related to the uncertainty of the applied current (instrumental error), which is on the order of 0.1 µA for a 20 µA applied current. We can therefore assume a constant proton flux within the instrumental parameters used, which is also reflected in the linearity of calibration curves that obey the Sand equation. In summary, indirect support for the constant nature of the proton flux is obtained by: i) the controlled current technique (J = i/nFA), ii) the high hydrogen ion concentration in the membrane, iii) no evidence of hydrogen ion depletion either at the inner or outer sides of the membrane during the current pulse by a lack of potential transitions, iv) excellent linearity in the τ1/2 vs concentration (Sand plot, see below), v) observation of the same baseline after depletion of base, indicating insignificant variation of potential with respect to time (see below).


12

Page 13 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2a-b shows the concentration profiles in the diffusion layer as a function of position for B and BH+ when a current of 20 µA [0.71 µA.mm-2] is applied across the membrane for 5-s (ci=0.5 mM, pKa=9). From the concentration profiles at the position 0 and using the Nernst equation, the expected chronopotentiogram is calculated. Figure 2c displays the simulated chronopotentiogram where four regions can be differentiated. The first part should be mainly dominated by a capacitive currents and ohmic drop (region I). The readout potential abruptly changes between the open circuit potential (OCP) to about 50-100 mV (for the experiment the potential change is about 400 mV) in a small fraction of time. Note that the model does not consider capacitive effects. The second part is related to the consumption of base in the diffusion layer. The concentration of the base changes as a function of time, resulting in a variation of pH but maintaining the same concentration gradient, see Figure 2a. Only a slight increase of potential is observed (region II). At the transition time, the concentration of the base is reduced to almost zero and an inflection of the potential–time trace is obtained. After this time (cB∼ 0, for x=0), the concentration of hydrogen ions rapidly increases (region III). A clear transition is observed in this particular case (see Figure 1S), but the sharpness of the transition is dictated by the pH change from before to after the transition, which is related to the initial concentration and the pKa of the base (see below). Finally, to maintain the applied current across the interface, the proton concentration gradually increases until the end of the experiment (region IV). The discussion above is illustrated with Figure 3 where the concentration profiles of all involved species in the flash titration are simulated. The transition time is here reached at 2.6-s. Note that the concentration of free hydrogen ions before the transition time is small (10-8 M) and therefore not appreciable in the speciation graph. After the transition time, the hydrogen ion concentration increases considerably up to 0.2 mM. Simultaneously, the protonated base also


13


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 32

reaches the level of the initial concentration (see Figure 2S, proton concentration profile after transition time). Several weak bases such as monohydrogenphosphate (HPO42-), ammonia (NH3), tris(hydroxymethyl)-aminomethane) (Tris), borate (B(OH)4-) and carbonate (CO32-) were measured at the same current density (20 µA, 0.7 µA.mm-2; see experimental section for sample preparation). Experimental chronopotentiograms as a function of the different non-protonated base concentrations are shown in Figure 4. All recorded chronopotentiograms show the four different regions mentioned above, however, the potential changes at the transition time are less sharp than predicted by the model (see Figure 3S, predicted chronopotentiogram for different concentrations). This is likely because of enhanced diffusion at the membrane edge where the assumption of one dimensional diffusion is not strictly valid. Nonetheless, the transition time is easily visualized as the maximum of the time derivative of the potential (Figure 5) and a precise transition time can be extracted from each processed chronopotentiogram. Even if the peak is broad as a result of a small pH difference (i.e, Figure 5a, HPO42-), the transition time is still reliable extracted. According to the Sand equation (Eqn. 5), a linear relationship should be found between the square root of the transition time and the bulk concentration of the non-protonated base. Figure 6 shows the calibration curves for six bases as a function of the concentration (left y-axis). In order to demonstrate that this concept works satisfactorily in a defined concentration range, the pH of the solution was alternately measured (before and after galvanostatic interrogation, solutions are stirred for 2 min to ensure a homogeneous concentration in the solution) by an external pH electrode as well as using the same perm-selective membrane used in chronopotentiometry. No significant difference in the pH determination was found between the commercial pH and the


14

Page 15 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


membrane under study (data not shown). The pH of the solution is plotted in Figure 6 (right yaxis). While the pH is practically constant during the experiment (pH ± 0.1) for all the bases except hydroxide, a linear relationship is observed in chronopotentiometry (τ0.5 vs concentration), suggesting the local titration of the base. As recently reported, the Sand equation can not only be used as an analytical tool to predict an unknown concentration, but can also be useful in determining the diffusion coefficient of the depleted species.22 Considering the geometry of the membrane, the reported diffusion coefficients29-32 and the remaining variables in Eqn. 5, the square root of transition times are calculated for each concentration. Figure 7 shows the experimental data (symbols) together with the predicted data (line) by the Sand equation. Each concentration point was repeated three times using four different membrane electrodes (RSD, 2%). Even though the predicted behavior fit well with theory, some deviations are detectable at low concentrations, particularly for Tris and borate. While the experimental data for ammonia follows a linear relationship with concentration, the data do not correspond quantitatively with the expected behavior from the Sand equation, owing to differences in diffusion coefficients (a 100% discrepancy between experimental data and theory is observed). A deposition of a hydrogel on the top of the membrane is expected to help reduce some undesired effects from the sample phase such as convection, temperature gradients and changes in viscosity. Table 1 summarizes the key analytical parameters of Figure 7 and compares experimentally observed diffusion coefficients with literature values. In order to understand the possible limitations of the proposed methodology, we simulated the titrimetric analysis using several bases with different pKa values (for simplification, monoprotonated bases were selected). Using Eqn. 10 and Eqn. 12 with their respective boundary


15


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 32

conditions, we calculated the free concentration of hydrogen ions at the membrane surface during a chronopotentiometric experiment. The sharpness of the transition time is reduced importantly with decreasing pKa (Figure 8a). In fact, a pKa of 7-7.5 may be currently estimated as the lowest limit for a concentration of 0.5 mM. Indeed, when weak bases such as MES (pKa 6.1) or acetate (pKa 4.75) were measured using flash-chronopotentiometry, no transition time was observable. Note that the limitations for numerically simulated chronopotentiometry are better, and similar to that of classical bulk titrations (see Figure 8b) at the same concentration (0.5 mM). As noted above, this is mainly due to a lower than expected sharpness of the transition observed experimentally. Because of the above stated limitation, we cannot experimentally observe the conversion of hydrogen carbonate to carbonic acid. However, as the initial sample contains around 100 times more HCO3- than CO32- (pH=8.7), the converted fraction of H2CO3 is relatively small compared to the remained HCO3- concentration in the time scale of our experiments. As a result, we report on the titratable concentration of carbonate, which is understood as P-alkalinity instead of total alkalinity. In contrast to the carbonate system, we explored a sample that contains sodium phosphate (pH=12.6). Two transition times were observed in chronopotentiometry, indicating the depletion of phosphate and monoacid phosphate is their expected order (see Fig 4S in the Supporting Information). A determination of P-alkalinity was performed in a sample from the river Arve in Geneva (pH=8.24). Figure 9 shows the observed chronopotentiograms for the sample (on the left) and for the standard additions of carbonate (on the right). The predicted P-alkalinity was 0.13 ± 0.01 mM, using the standard addition calibration (τ1/2 = 0.31 s1/2 mM-1 + 2.28 s1/2). An acid titration8 with phenolphthalein indicator gave a P-alkalinity level of the sample of 0.14 ± 0.01 mM (SD, n


16

Page 17 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


= 5), which is in quantitative agreement with the obtained carbonate concentration by the chronopotentiometric method.


17


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 32

CONCLUSIONS In summary, we presented here a fundamental study of a confined flash-chronopotentiometry titration in the vicinity of the surface of a permselective membrane. We demonstrated the capacity of such membranes to measure pH and alkalinity sequentially of artificial samples. A model that explains the observations was developed. In addition, numerical simulations were used to overcome the differential equation that represents the diffusion of several ions in the course of the experiment. This resulted in the computation of ideal chronopotentiograms as a function of the base concentration. The predicted and the obtained chronopotentiogram show comparable behavior, although experimental transition times are not as sharp as predicted. Several bases with different pKa values were satisfactorily measured between 0.1 to 1.0 mM. As a quality parameter of our data, the diffusion coefficients were extracted from the calibration curves. Such diffusion coefficients correlated well with the reported values in most of the cases. In view of understanding the limitations of the proposed system, we studied the influence of the pKa on the visibility of the transition time. Bases with a pKa lower than 7.5 are not satisfactorily detected at less than millimolar concentrations. It is noted that this pKa limit is important for environmental real samples where the initial pH is in the range of 5 to 8. While beyond the scope of this work, future efforts will need to focus on the performance of this method in environmental and other important samples. As an example of the analytical potential of this methodology, P-alkalinity of a river sample was found to be in agreement with titrimetric analysis. It is our aim to eventually incorporate such alkalinity sensors in submersible sensing probes to achieve the measurement of concentration profiles as a function of depth in natural freshwater and oceanic systems.


18

Page 19 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ASSOCIATED CONTENT Supporting Information. AUTHOR INFORMATION Corresponding Author *[email protected] ACKNOWLEDGMENT The authors thank the Swiss National Science Foundation and the European Union (FP7-GA 614002-SCHeMA project) for supporting this research.


19


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 32

REFERENCES (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)

Minich, D. M., Bland, J. S., Altern Ther Health Med 2007, 13. 62-65 Udeigwe, T. K., Eze, P. N., Teboh, J. M., Stietiya, M. H., Environ. Int. 2011, 37. 258-267 Bakker, E., Buhlmann, P., Pretsch, E., Chem. Rev. 1997, 97. 3083-3132 Belyustin, A. A., J. Solid State Electrochem. 2011, 15. 47-65 Mikhelson, K. N., Ion-Selective Electrodes. Berlin Heidelberg: 2013. Asuero, A. G., Michalowski, T., Crit. Rev. Anal. Chem. 2011, 41. 151-187 Dickson, A. G., Deep-Sea Research Part a-Oceanographic Research Papers 1981, 28. 609-623 Michalowski, T., Asuero, A. G., Crit. Rev. Anal. Chem. 2012, 42. 220-244 van der Schoot, B., van der Wal, P., de Rooij, N., West, S., Sens. Actuators B Chem. 2005, 105. 88-95 Ciszkowska, M., Stojek, Z., Morris, S. E., Osteryoung, J. G., Anal. Chem. 1992, 64. 2372-2377 Daniele, S., Bragato, C., Baldo, M. A., Electrochim. Acta. 2006, 52. 54-61 Daniele, S., Bragato, C., Baldo, M. A., Mori, G., Giannetto, M., Anal. Chim. Acta 2001, 432. 2737 Jaworski, A., Osteryoung, J. G., Donten, M., Stojek, Z., Anal. Chem. 1999, 71. 3853-3861 Roberts, J. M., Linse, P., Osteryoung, J. G., Langmuir 1998, 14. 204-213 Stojek, Z., Ciszkowska, M., Osteryoung, J. G., Anal. Chem. 1994, 66. 1507-1512 Wen, X. W., Herdan, J., West, S., Kinkade, D., Vilissova, N., Anderson, M., J. AOAC Int. 2004, 87. 1208-1217 Horvai, G., Pungor, E., Anal. Chim. Acta 1991, 243. 55-59 Hanselman, R. B., Rogers, L. B., Anal. Chem. 1960, 32. 1240-1245 Bhakthavatsalam, V., Shvarev, A., Bakker, E., Analyst 2006, 131. 895-900 Bhakthavatsalam, V., Bakker, E., Electroanalysis 2008, 20. 225-232 Gemene, K. L., Bakker, E., Anal. Chem. 2008, 80. 3743-3750 Afshar, M. G., Crespo, G. A., Bakker, E., Anal. Chem. 2012, 84. 8813-8821 Crespo, G. A., Afshar, M. G., Bakker, E., Angew. Chem. Int. Ed. Engl. 2012, 51. 12575-12578 Crespo, G. A., Afshar, M. G., Bakker, E., Anal. Chem. 2012, 84. 10165-10169 Crespo, G. A., Bakker, E., Rsc Advances 2013, 3. 25461-25474 Buhlmann, P., Pretsch, E., Bakker, E., Chem. Rev. 1998, 98. 1593-1687 Grygolowicz-Pawlak, E., Crespo, G. A., Afshar, M. G., Mistlberger, G., Bakker, E., Anal. Chem. 2013, 85. 6208-6212 Morf, W. E., Pretsch, E., de Rooij, N. F., J. Electroanal. Chem. 2010, 641. 45-56 Bard, A. J., Anal. Chem. 1961, 33. 11-& Buffle, J., Zhang, Z., Startchev, K., Environ. Sci. Technol. 2007, 41. 7609-7620 Ng, B., Barry, P. H., J. Neurosci. Methods 1995, 56. 37-41 Frank, M. J. W., Kuipers, J. A. M., vanSwaaij, W. P. M., J. Chem. Eng. Data 1996, 41. 297-302


20

Page 21 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figures Captions Figure 1.

Schematic illustration of current-driven proton release for alkalinity determination. I) Interrogation step (anodic current), II) Regeneration step (applied potential at the previously determined open circuit potential). B (base), BH+ (protonated base), RE (reference electrode), WE (working electrode), CE (counter electrode), L (chromoionophore I), R- (cation exchanger), A (current), V (potential), IFS (internal filling solution).

Figure 2.

Concentration profiles in the diffusion layer as a function of time for a) BH+ and b) B during 5-s of a chronopotentiometric experiment. Position 0 corresponds to the sample-membrane interface. c) Corresponding chronopotentiogram with its respective regions as described in the text. Parameters: ci=0.5 mM, pKa=9, A=28.2 mm2, Δt=1ms, DB=1 10-5 cm2s-1, i=20 µA, d=5.10-4 cm.

Figure 3.

Simulated concentration changes for the indicated species at the membrane-sample interface (position 0) during the chronopotentiometric experiment (5-s). Used parameters: ci=0.5 mM, pKa=9, A=28.2 mm2, Δt=1ms, DB=1 10-5 cm2s-1, i=20 µA, d=5.10-4 cm.

Figure 4.

Experimental chronopotentiograms for the detection of a) monohydrogenphospate, b) ammonia, c) Tris, d) hydroxide, e) borate and f) carbonate. The base concentration is given in units of millimolar.

Figure 5.

Time derivatives of the chronopotentiograms given in Figure 4 with indicated millimolar base concentrations.

Figure 6.

Potentiometric (pH, see right y-axis) and chronopotentiometric (square root of transition time, left y-axis) data for a) monohydrogenphosphate, b) ammonia, c) Tris, d) hydroxide, e) borate and f) carbonate. Straight lines for the chronopotentiometric data are least squares fits.

Figure 7.

Calibration curves give as chronopotentiometric signal readout (square root of time) vs. base concentration. Each data set (markers) is compared with a straight line described by the Sand equation and using a reported diffusion coefficients. Only the experimental data for ammonia do not correspond quantitatively to the expected diffusion coefficients, see Table 1.

Figure 8.

Influence of the pKa values on simulated chronopotentiograms a) and volumetric titrimetric analysis b). The pKa values are given on the graph. Other parameters: ci=0.5 mM, A=28.2 mm2, Δt=1ms, DB=1 10-5 cm2s-1, i=20 µA, d=5.10-4 cm.

Figure 9.

Obtained chronopotentiograms in a sample of the river Arve in Geneva (pH=8.24). The first peak on the very left corresponds to the river sample without carbonate addition. The remaining peaks to the right correspond to standard additions of carbonate. Inset: square root of time as a function of carbonate concentration.


21


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 32

Figure 1


22

Page 23 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2


23


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 32

Figure 3


24

Page 25 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4


25


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 32

Figure 5


26

Page 27 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 6


27


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 32

Figure 7


28

Page 29 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 8


29


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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 32

Figure 9


30

Page 31 of 32

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1. Diffusion coefficients, intercepts and linear ranges extracted from Figure 7 for all the measured bases. Base (pKa)

Expected DB30-32 / cm2.s-1

Experimental DB / cm2.s-1

Intercept / s-1/2

Linear range / mM

HPO42- (7)

7.51 10-6

(7.6±0.2) 10-6

0.07±0.03

0.1-1.0

NH3 (9.2)

1.51 10-5

(8.1±0.3) 10-6

0.10±0.02

0.1-1.2

Tris (8.1)

7.2 10-6

(7.1±0.1) 10-6

0.15±0.04

0.1-1.0

OH-

5.27 10-5

(5.2±0.1)10-5

0.15±0.04

0.2-0.4

B(OH)4- (9)

---

(1.2±0.3)10-5

0.07±0.03

0.1-0.6

CO32- (10.3)

9.22 10-6

(9.1±0.4) 10-6

-0.05±0.03

0.2-1.2


31


For TOC only:

MEMBRANE SAMPLE

R– IndH+

B BH+

APPLIED CURRENT

POTENTIAL

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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 32

TIME


32

Direct Alkalinity Detection with Ion-Selective Chronopotentiometry

Recommend Documents