~2S733~
SENSOR HAVING ION-SELECTIVE ELECTRODES
Description
Background
Ion-Selective Electrodes
Ion-selective electrodes respond preferentially
or selectively to a particular ionic species in a
liquid. They are often used in potentiometric
measurement of the activity of an ion in a liquid
sample. Potentiometric measurement determines the
difference in electrical potential between two
electrodes which, in contact with a liquid, form an
electrochemical cell. The half cell potential of
one electrode - the reference electrode - is essen-
tiaîly constant and that of the other electrode -
the indicator electrode - varies with the ionic
activity of the liquid being analyzed. The elec-
trical potential across the electrodes is propor-
tional to the logarithm of the activity of ions in
solution to which the ion-selective electrode
responds. The Nernst equation describes the loga-
rithmic relationship. The difference in electrical
potential can be determined using a potentiometric
measuring device, such as an electrometer.
Several types of ion-selective electrodes are
available and include, for example, conventional
1~;7~3.'1
glass electrodes for pH determinations which are
widely used in laboratories. Glass electrodes are
based on alkali ion silicate compositions. Elec-
trodes for the determination of pH can be made of
lithium silicates or borosilicate glass which is
permeable to hydrogen ions (H~) but not to anions or
to other cations. If a thin layer of a glass
selectively permeable to H is positioned between
two solutions of different H concentrations, H
ions will move across the glass from the solution of
high concentration to that of low H concentration.
Such movement results in the addition of a positive
ion to the solution of low H concentration, leaves
a negative ion on the opposite side of the glass and
produces an electric potential across the glass.
Immersion of a pH independent reference electrode
into the solution completes the circuit and the
potential difference can be measured.
Another type of ion-selective electrode uses
,20 liquid ion exchangers and is supported by inert
polymers, such as cellulose acetate, or in polyvinyl
chloride films. An important example of this type
of electrode is a Ca -responsive electrode which is
based on calcium salts of diesters of oil-soluble
phosphoric acids. U.S. Patents 3,429,785;
3,438,886; and 3,445,365 describe ion-sensitive
electrodes which have membranes made of a porous
inert substance filled with an ion-exchange organic
liquid, which are selectively responsive to divalent
cations (e.g., Ca , Mg ).
~2~;733~
--3--
Neutral carrier-based sensors for both mono-
s valent and divalent cations are similar to ion-
exchanger-based electrodes. soth can involve
ion-exchange sites, especially negative mobile sites
5 resulting from mediators or negative fixed sites
arising from hydrolysis of support materials.
Neutral carriers, which can be cyclic or open-chain,
are generally hydrophobic complex formers with
cations. Such compounds result in selective extrac-
10 tion, and therefore selective permeability, for K ,
Na and Ca2~, which would not otherwise dissolve in
the membrane phase as simple inorganic salts. An
example of such a carrier is valinomycin, which can
be used in electrodes selectively responsive to K .
Electrodes for the determination of the H
content of a liquid sample have been described by
others. They commonly contain a plastic membrane
which has an ion-selective component (an ionophore)
and a solvent/plasticizer compound in which the
20 ion-selective component can be dissolved. In
addition to glasses selectively permeable to H
ions, hydrogen ionophores such as lipophilic deri-
vatives of uncouplers of oxidative phosphorylation
and lipophilic tertiary amines have been used.
Simon et al, for example, have described a
micro pH electrode containing tridodecylamine
(C12H25)3N or trilauryl amine) as the hydrogen
ionophore. Analytical Chemistry 53: 2267-2270
(1981). The authors report that this electrode
gives excellent response to changes in pH between
5.5 and 12. However, at least two important
;
--4--
disadvantages result from the fact that the membrane
phase must contain dissolved carbon dioxide (CO2) in
order to function properly. First, electrode
performance (i.e., CO2 content) may be seriously
affected by changes in temperature and pressure.
Second, mass production is also very difficult
because manufacture must be carried out under
carefully controlled conditions; in addition, the
dissolved CO2 diffuses out during storage of the
electrodes.
Other types of ionophores are based on un-
couplers of oxidative phosphorylation in mito-
chondria. An example is described by Finkelstein.
Biochimica BioPhYsica Acta, 205: 1-6 (1970).
Weak-acid uncouplers of phosphorylation, such as
2,4-dinitrophenol and m-chlorophenylhydrazone
mesoxalonitrile, act as H ion carriers. They ar~,
however, unsuitable as components of membranes
incorporated into ion-selective electrodes because
of their finite water solubilities (i.e.~ they would
not remain membrane bound and would leach out of the
membrane).
Brown et al. describe a pH sensor claimed to be
suitable for chronic intravascular implantation.
The pH-sensitive element is a thin film of an
elastomeric polymer which is made ion permselective
through the addition of a hydrophobic, lipophilic
specific H -ion carrier. The carrier used is
p-octadecyloxy-m-chlorophenylhydrazone
mesoxalonitrile tOCPHr which is a higher molecular
weight homolog of the weak-acid uncoupler
~2~;73:~
--5--
m-chlorophenylhydrazGne mesoxalonitrile). Brown and
- co-workers state that although the OCPH molecule
acted as a mobile H carrier in a variety of elas-
tomers, pH response characteristics of sufficiently
good quality for practical use were only obtained
using polymer matrices especially designed for the
purpose. U.S. Patent No. 3,743,588 (1973); O.H.
LeBlanc, J.F. Brown et al. Journal of Applied
PhYsiology, 40: 644-647 tl976). The preparation of
polymers which can be used with the OCPH molecule is
described by Vaughn. U.S. Patent No. 3,189,662
(1965); H.A. Vaughn and E.P. Goldberg, Polymer
Letters, 7: 569-572 (1969).
In U.S. Patent 3,691,047 (1972) Ross and Martin
describe an ion-sensitive membrane for potenti-
ometric electrodes. The membrane is described as a
gelled mixture in which the solid phase is described
as being polymer (e.g., cellulose triacetate) and
the liquid phase as an organic ion-exchange material
dissolved in an organic solvent (e.g., calcium
(bis-dibromoleylphosphate)2 dissolved in dioctyl-
phenylphosphonate). The ion-sensitive electrode is
said to have a membrane whose major constituent is
an organic ion-exchanger which is dissolved in a
non-volatile organic solvent.
In U.S. Patent 4,214,968 (1980), Battaglia et
al. describe a dry-operative ion-selective electrode
for use in determining ion content of liquids. The
electrode is said to be comprised of a dry internal
reference electrode in contact with a hydrophobic
ion-selective membrane. The internal reference
--6--
electrode is a dried metal/metal-salt reference half
cell or a dried redox couple reference electrode and
is wetted upon application of a liquid sample. The
ion-selective membrane includes an ion carrier
(e.g., valinomycin) which is dissolved in a carrier
solvent dispersed in a hydrophobic binder.
In U.S. Patent 4,184,936 (1980), Paul and
sabaoglu describe a device for determining ion
activity of a liquid sample. The device is de-
scribed as having an ion-selective membrane which is
coated over an internal reference element (made of
an electrolyte-containing layer, a metal salt and a
metal layer) and a support. The two electrodes of
the device are said to be solid and, preferably,
dried.
In U.S. Patent 4,053,381 (1977), Hamblen et al.
describe another device for measuring ion activity
in liquids. The electrodes which are a component of
the device preferably include at least one ion-
selective electrode in which the internal referenceelectrode has several layers. The layers include a
metal layer, a layer of an insoluble salt of the
metal and an electrolyte-containing layer which is
preferably dried. The claimed device includes solid
electrodes.
At the present time, there are many types of
ion-selective electrodes available for the measure-
ment of the ion content of a liquid. These ion-
selective electrodes, however, have limitations.
These limitations include the requirement for
membranes comprised of specially designed polymer
. .
~S~
--7--
matrices; utilization of ionophores which require
pre-neutralization with base to improve membrane
sensitivity and reduce response time; the need for
storage under well-controlled conditions; and loss
of sensitivity and reliability during storage. In
addition, ion-selective electrodes now available for
pH determination require relatively large samples
(i.e., 1.0 ml. or more) for accurate operation and
are made of glass, which is costly and cannot be
incorporated into an electrode suitable for automa-
tic processing of samples of very small size.
There is a need for an ion-selective electrode
which can be used to provide accurate, rapid
measurement of the ion content of smaller samples
(e.g., of microliter size) than those presently
available. There is also a need for an ion-
selective electrode which can provide an accurate
and quicker measurement of other constituents of
smaller samples than is possible with presently
available electrodes.
Differential Measurement Techniques
The technique of differential potentiometric
measurement depends on potential differences arising
between two identical electrochemical half-cells
immersed in solutions of different activity sepa-
rated by a salt bridge. The two half-cells together
comprise a concentration cell. In the present case
the activity of one half-cell (a1) is fixed
(reference) while that of the other (a2) (sample) is
variable such that the emf of the concentration cell
may be defined as:
~2S733~
--8--
1. El = E ~ RnF lnal
RT
2. E2 E nF 1 2
Ecell E2-El = nF lna2 ~ nF lna
4- Ecell = nF ln al
where RnT = 59.1 mv @ 298K for n=l
and,
El = emf of reference half-cell
E2 = emf of sample half-cell
E - standard reference electrode potential
R = molar gas constant; 8.314
volt-coulombs/degree(K)
T = absolute temperature, K
n = charge on the ion
F = Faraday constant; 96,493 coulmbs
al and a2 = ionic activity of reference sample
In the case of a cation such as potassium (K ), if
the half-cell al is defined as a reference half-cell
and assigned a zero potential then the emf of the
concentration cell will be positive if a2~ al and
negative if a2 <al. Since al is fixed, the equation
for the cell potential contains only one unknown
(a2); upon measuring ECell the equation may be
solved for a2.
Differential measurement techniques have been
used to measure indirectly the concentration or
activity of constituents of biological fluids other
than H , including other ions such as sodium (Na ),
potassium (K ), calcium (Ca ) and chloride (Cl ).
In addition, such techniques often make use of
3~2S~7331
biosensors or enzyme electrodes, which include a
biological catalyst (e.g., immobilized enzymes,
cells, layers of tissue) coupled to an electrode
sensitive to a product or co-substrate of the
biologically catalyæed reaction. The concentration
of enzymes or of substrates can be determined using
differential measurement techniques. For example,
many enzyme reactions result in the production of an
acid or a base. Ionization of the acid or base in
turn results in liberation or uptake of H and a
change in the pH of the solution. The measured
change in H concentration or pH can be the basis
for a stoichiometric determination of the concen-
tration of substances le.g., glucose, urea, etc.)
which liberate or take up hydrogen ions.
In the case of differential pH measurement,
each half-cell would contain a pH electrode. The
hydrogen ion activity of the al half-cell would be
fixed and that of a2 ~sample) would vary depending
on the pH of the sample. The emf measured across
the cell by means of an electrometer could then be
used to calculate the hydrogen ion activity of a2.
In the case of a differential measurement
requiring an enzyme, the enzyme would be placed in
either one or both half-cells and the sample would
be added to both half-cells or the sample would be
added to one half-cell and a calibrator would be
added to the other half-cell. As a result, when the
enzyme reacts with the substrate in the sample, a
decrease or increase in pH occurs; the magnitude of
this change is directly proportional to the amount
~2~i7331
--10--
of substrate in the sample. Similarly, a substrate
- could be substituted for the enzyme and the cell
used to measure the enæyme activity of a given
sample.
For example, Nilsson and co-workers describe
the development of enzyme electrodes in which
hydrogen ion glass electrodes are used to make
enzyme-pH electrodes for the determination of
glucose, urea and penicillin in solutions. The
enzymes used for the determination are glucose
oxidase, urease and penicillinase, respectively.
Nilsson, H. et al., Biochimica et Biophysica Acta,
320:529-534 (1973).
Mosca and co-workers also describe the determ-
ination of glucose by means of differential pH
measurements. The technique is based on the mea-
surement of the change in pH produced by the hexo-
kinase catalyzed reaction between glucose and ATP.
They describe two systems said to be useful for
determining the difference in pH between two l-ml.
aqueous samples. The concentration of glucose is
calculated from the measured change in pH by means
of an equation derived by the authors. Mosca, A. et
al., Analytical Biochemistry, 112:287-294 (1981).
Differential measurement of pH to determine glucose
in whole blood and plasma and development of an
automated system for doing so is subsequently
described by this group. Luzzana, M. et al.,
Clinical Chemistry, 29:80-85 (1983).
The same apparatus and a differential pH
technique are described for use in measuring lipase
12S7331
--11--
activity of biological fluids such as serum, plasma
and duodenal juice. Ceriotta, F. et al., Clinical
Chemistry, 3I:257-260 (1985). A refinement in the
differential pH measurement technique is said to
serve as the basis for the determination of urea,
creatinine and glucose in plasma and whole blood.
The enzymes urease, creatinine iminohydrolase and
hexokinase, respectively, are used for the determin-
ations. The concentrations of the three substrates
are calculated based on the observed changes in pH
of the solutions after reactions have occurred.
In U.S. Patent 4,353,867 (1982), Luzzana
describes a method and apparatus for the determin-
ation of substances, such as glucose, urea and
enzymes in biological solutions (e.g., blood, serum,
urine). The method uses differential pH measurement
in which two glass pH electrodes are placed in
separate solutions; the change of pH in the solu-
tions after reagents are added is determined; and
the concentration of the substance of interest is
calculated from the observed pH change. The appa-
ratus is comprised of a sample cuvette; a cell
having two glass capillary electrodes; a means for
measuring pH at the two electrodes; and an elec-
tronic means for calculating the concentration ofthe substance from the pH measurements.
Immunosensors
Electrochemical immunosensors may be described
as either potentiometric or amperometric. Potenti-
ometric immunosensors may be used to measure either
~25~331
antibodies or antigens. They may be described aseither direct or indirect and can be either mem-
branes or solid electrodes. An example of a direct
potentiometric immunosensor for the determination of
an antigen is described by Yamamoto, et al. An
antibody, anti-hCG, is immobilized on a titanium
wire. The anti-hCG electrode and a reference elec-
trode are placed in a buffer solution. The antigen,
hCG, is added and as the antigen binds to the
immobilized antibody, the potential difference
between the two electrodes changes until equilibrium
is attained. The equilibrium potential difference
is directly proportional to the concentration of the
antigen. Yamamoto, et al., Clinical Chemistry, 26:
1569-1572 (1980). The exact nature of the potenti-
ometric response is not fully understood, but it is
- generally acknowledged to involve a surface charge
neutralization or redistribution.
Another example of a direct potentiometric
immunosensor for the determination of an antibody is
described by Keating and Rechnitz. This type of
immunosensor responds to specific antibodies through
modulation of a background potential, fixed by a
marker ion, in such a manner that the potential
change is proportional to the concentration of
antibody. An immunosensor for the determination of
digoxin antibody is described wherein digoxin is
coupled to the marker ion carrier molecule benzo-
15-crown-5. The resulting conjugate is incorporated
into a polyvinylchloride membrane. Keating, M.Y.
and Rechnitz, G., Analytical Chemistry, 56: 801-806
, '
.
~:
1~5733~
-13-
(1984). An antibody-selective potentiometric elec-
trode for antibody determination is also described
by Rechnitz and Solsky in U.S. Patent 4,402,819. A
serious limitation to the routine clinical use of
this method is the necessity of maintaining a
constant background level of the marker ion such
that biological samples would have to by dialyzed
before analysis.
Indirect potentiometric immunosensors are
enzyme linked and are analogous to other enzyme
immunoassay methods except that the electrochemical
sensor is used to measure the product of the enzyme-
substrate reaction. Both homogeneous and hetero-
geneous potentiometric enzyme immunoassays have been
described. For example, Boitieux and coworkers
describe a heterogeneous potentiometric enzyme-
linked immunoassay for the determination of
estradiol. Boitieux et al., Clinical Chemica Acta,
113: 175-182 (1981). The antibody to estradiol is
immobilized onto a porous gelatin membrane. The
membrane is incubated with peroxidase-labelled
estradiol and free estradiol. After washing, the
membrane is fixed onto an iodide sensitive elec-
trode. The peroxidase activity is determined in the
presence of hydrogen peroxide and iodide ion. The
iodide selective electrode potential is a function
of the estradiol concentration.
A homogenous potentiometric enæyme immuno-
assay for human IgG has been described by Fonong and
Rechnitz. The method is based on the inhibition, by
IgG, of CO2 production by beta-ketoadipic acid
125733~
-14-
catalyzed by chloroperoxidase enzyme conjugated to
- IgG antibody. Fonong and Rechnitz, G., Analytical
Chemistry, 56: 2586-2590 (1984). If the enzyme-IgG
conjugate is incubated with a sample containing the
antigen (IgG) before reaction with the enzyme
substrate, the observed rate of CO2 liberation,
measured with a potentiometric CO2 gas-sensing
electrode, will be decreased. The decrease in
activity is proportional to the concentration of IgG
in the sample.
Amperometric enzyme immunosensors are analogous
to the potentiometric immunosensors except that an
amperometric sensor, usually an oxygen electrode, is
used to measure enzyme activity. For example,
Aizawa, Morioko and Suzuki have described an amper-
ometric immunosensor for the determination of the
tumor antigen alpha-fetoprotein (AFP). Aizawa et
al., Analytica Chimica Acta, 15: 61-67 (1980).
Anti-AFP is covalently immobilized on a porous
membrane. The membrane is incubated with catalase-
labelled AFP and free AFP. After competitive
binding the membrane is examined for catalase
activity by amperometric measurement of oxygen after
addition of hydrogen peroxide. In a similar manner,
2S Boitieux and co-workers describe an amperometric
enzyme immunoassay for the determination of hepa-
titis B surface antigen. Boitieux et al., Clinical
Chemistry, 25: 318-321 (1979).
A limitation of enzyme-linked amperometric
sensors is the necessity of synthesizing the enzyme
~2~i~331
-15-
eonjugates and the high eost of currently available
amperometrie sensors sueh as oxygen electrodes.
Disclosure of the Invention
-
The present invention is a sensor for the
potentiometric determination of the ion content or
activity of a sample and the concentration of other
components of a sample through the use of ion-
selective electrodes. The sensor is especially
suited for rapid determination of the hydrogen ion
content or pH of biological fluids; the concentra-
tion of other ions in biological fluids; and,
through differential pH measurements or immunoassay
techniques, the concentration of other components
(e.g., glucose, urea, triglycerides, uric acid
enzymes such as aspartate aminotransferase (AST),
alanine aminotransferase (ALT), amylase, creatinine
kinase (CK), alkaline phosphatase and drugs) of bio-
logical fluids. It is particularly useful for
automated handling or processing.
The sensor is comprised of ion-selective
electrodes whieh are held in a frame and have porous
material (e.g., mieroporous polyethylene sueh as
that manufaetured by Porex Teehnologies Corp.)
between them. The porous material provides a means
for ionie flow between the eleetrodes onee a sample
is applied at the electrodes. The ion-selective
eleetrodes are eomprised of a membrane which is
seleetively permeable to the ion or other substance
whose eoneentration is to be determined and a
~2S733~
-16-
reference electrode. The membrane does not require
- preconditioning before use.
The selectively permeable membrane is comprised
of an ion-selective compound (or ionophore) and a
thermoplastic resin or a plastic material, which can
all be dissolved in an organic solvent. In addi-
tion, it can also include a plasticizer. The
ion-selective component for a pH sensor is a com-
pound having the following general formula:
Q
~ C/~
wherein
Rl, R2 and R3 are independently selected andeach can be: 1) a halogen: 2) an alkyl group having
from 4-18 carbon atoms; 3) a halogen-substituted
alkyl group; 4) an alkoxy group; 5) a halogen-
substituted alkoxy group; 6) an acid group repre-
sented by -CO2R4 wherein R4 is alkyl having 1-18
carbons; 7) a keto group represented by -COR5
wherein R5 is selected from the groups defined for
R4; or 8) a hydrogen atom.
In an embodiment of the present invention, the
membrane comprises an organic plastic matrix,
polyvinyl chloride (PVC), which contains the ion-
selective compound 2-octadecyloxy-5-carbethoxy-
phenylhydrazone mesoxalonitrile. The ion-selective
compound, the PVC and a plasticizer, which in this
embodiment is 2-nitrophenyloctylether, are all
~25733~
-17-
soluble in the solvent tetrahydrofuran (THF). These
components comprise the membrane formulation.
In a preferred embodiment, the membrane com-
prises about 10-40% PVC by weight, and its thickness
is greater than 1 mil. In a particularly preferred
embodiment, the membrane is made of about 20-35~ PVC
by weight and is about 3-15 mils thick. The mem-
brane which is described can be used as a component
of the ion-selective electrode of the present
invention. It can also be incorporated into
commercially available electrode bodies.
The sensor of this invention also has an
internal reference element or material which in-
cludes a known concentration of the sample component
whose concentration is to be determined.
Brief Description of the Drawings
Figure 1 is a perspective view of the top of a
sensor having ion-selective electrodes and a handle
which has a magnetic stripe in which data is
recorded.
Figure 2 is a perspective view of the bottom of
a sensor having ion-selective electrodes and a
handle which has a magnetic stripe in which data is
recorded.
Figure 3 is a perspective view showing the
individual components of a sensor which can be used
to determine the hydrogen ion activity (pH) of a
sample or the concentration of other ions in a
sample.
.
.
~257~
-18-
Figure 4 presents schematic representations of
ion selective electrode configurations.
Figure 5 shows an ion-selective membrane of
this invention inserted into a commercially avail-
able barrel-type electrode.
Figure 6 is a perspective view showing the
individual components of a sensor which can be used
to determine the activity or concentration of a
component of a sample by a differential measurement
technique.
Best Mode for Carrying Out the Invention
The sensor which is the subject of this inven-
tion is used for the potentiometric determination of
the ion content of a sample or the concentration of
other components of a sample. It is particularly
useful in the rapid determination of the hydrogen
ion activity (or pH) of or concentration of other
ions in a biological fluid (e.g., blood) and in
measuring concentrations of other components (e.g.,
glucose, urea, triglycerides, uric acid, enzymes
such as aspartate aminotransferase (AST), alanine
aminotransferase (ALT), amylase, creatine kinase
(CK), alkaline phosphatase and drugs) in biological
fluids. Ionophores selective for sample components
and incorporated into ion-selective membranes of
such sensors are also the subject of this invention.
The sensor can now be further described with
reference to the figures.
Figures 1 through 3 show a sensor having
ion-selective electrodes and a handle having a
~257331
--19--
magnetic stripe in which data are recorded. The
sensor represented in these figures can be used to
determine the hydrogen ion content or pH of a sample
or the concentration of other ions in a sample.
These figures will now be referred to in describing
a sensor to be used for the determination of hydro-
gen ion content of a sample. It is to be under-
stood, however, that the sensor can be used for
determination of other ions or other components as
well.
The sensor 30 is comprised of ion-selective
electrodes which are held in a frame or body 10
which has upper section 12 and lower section 14. In
addition, the sensor 30 has a handle 5 bearing a
magnetic stripe 8 on its lower surface.
Upper section 12 of frame 10 has four openings
,7 4 and three grooves 6, which are positioned between
~ the openings. Lower section 14 of frame 10 has four
! ' openings 7 and three grooves (not shown) positioned
between the openings. Upper section 12 and lower
section 14 are in such a relationship that openings
4 and openings 7 are aligned and grooves 6 in the
upper section and the grooves in the lower section
are aligned. As a result, openings 4 and 7 form
four openings 11 and grooves 6 of the upper section
and the grooves in the lower section define spaces
between three of the openings 11. Upper section 12
has small holes 15 located between openings 4.
Holes 15 allow the passage of air. The ion-
selective electrodes are located within the openings11 and are connected by a porous material 17 which
~'257~3~
-20-
is cylindrical or rod shaped and provides a means
- for ionic flow between the electrodes upon the
application of a sample at the electrodes. The
cylindrical or rod-shaped porous material 17 is
located in the spaces between openings 11. Upper
section 12 and lower section 14 are sealed, for
example by being ultrasonically welded under pres-
sure, such that there is no leakage of sample or
solutions onto porous material 17 or between or into
the two sections of frame or body 10.
As shown in Figure 3, sensor 30 has four
positions at which ion-selective electrodes can be
located. Although there will generally be an
ion-selective electrode at each of these positions,
these can be used in various combinations depending
on the analytical information desired. Accordingly,
the position of grooves 6 in upper section 12 and
the grooves in lower section 14 will vary as needed
for the analysis being carried out. For example,
grooves 6 in upper secton 12 can be positioned as
shown in Figure 3; the grooves in lower secton 14
will be positioned so that spaces will be defined
between the three electrodes, as described above,
when upper section 12 and lower section 14 are
joined. Alternatively, two grooves 6 can be posi-
tioned in parallel to one another in upper section
12 and the grooves in lower section 14 positioned
correspondingly. Two, three or all four electrodes
can be used. Porous rods 17 can be hydrophilic and
conductive or hydrophobic and nonconductive. The
use of conductive or nonconductive material between
~Z57;~3~
-21-
a pair of electrodes is determined by the analyses
to be carried out. $he various combinations of
membranes are best described by example.
In the first example, as represented in Figure
4a, three membranes are active and all three contain
the same membrane. Positions a and b have known
concentrations o~ the ion to be determined and
position c has the sample, which has an unknown
concentration of the ion of interest. The emf
developed between the solutions in a and b can be
used to calibrate the sensor; the slope value is
then used to determine the concentration of the
sample from the emf developed between the solutions
in b and c. In addition, because the concentrations
of ions in a and b are known, it is possible, using
predetermined slope values, to calculate what the
potential difference between a and b should be and
to compare this value with the measured potential
difference or to calculate a concentration for a and
determine how this varies from the known value. In
a clinical chemistry sense, the solution in a would
then be used as a control.
The a, b, c configuration can also be used in
such a manner that replicate samples can be
analyzed. This is represented in Figure 4b. In
this case the same sample is placed in positions a
and c and reference solution in b. Using a pre-
determined slope value, it is possible to determine
simultaneously two values for the same sample.
It is also possible to determine simultaneously
two different analytes. This is represented in
~257~3~
-22-
Figure 4c. For instance, a and d can be membranes
- selective for one analyte and b and c can be mem-
branes selective for another analyte. Useful
combinations might be sodium/potassium, pH/calcium,
glucose/urea and urea/creatinine. In these in-
stances predetermined calibration data are required.
An enzyme/substrate sensor can be constructed
as represented in Figure 4d. In this case, a, b, c
and d can all be the same membrane; for example,
they can all be for determination of pH. However, a
and d would contain, in addition to the pH membrane,
an immobilized enzyme. An unknown analyte sample
concentration is added to both c and d; a known
concentration of the same analyte is added to a and
b. The enzymes in a and d act on the substrate and
produce a pH change proportional to the concentra-
tion of the analyte in the sample. If it is assumed
that all the membranes are identical, the emf
developed between a and b (the reference solution)
can be used to calibrate the electrodes. The
calibration data obtained for a and b may then be
used to calculate the sample concentration in c and
d.
An enzyme/substrate sensor can be constructed
as represented in Figure 4k. In this case, a, b and
c can all be the same membrane; for example, they
can all be for determination of pH. However, a, b
and c contain, in addition to the pH membrane, an
immobilized substrate. The substrate concentrations
in a and b are different; the concentration of
substrate in c can be the same as in a and b or can
~257~
-23-
be different. An unknown analyte enzyme sample
concentration is added to c; a known concentration
of the enzyme to be measured is added to a and b.
The enzymes in a, b and c act on the substrate in a,
b and c and produce a pH change proportional to the
concentration of the enzyme in the sample. If it is
assumed that all the membranes are identical, the
emf developed between a and b tthe reference solu-
tion) can be used to calibrate the electrodes. The
calibration data obtained for a and b may then be
used to calculate the enzyme sample concentration in
c.
Examples of substrates which can be immobilized
in an enzyme/substrate sensor such as that repre-
sented in Figure 4k are: amino acids such asL-alanine and L-aspartate, creatine and starch.
These can be used, respectively, for the determina-
tion in a sample of: alanine aminotransferase,
aspartate aminotransferase, creatine kinase and
amylase. As described in Example 8, an enzyme/
substrate such as that represented in Figure 4k can
also be used to determine the creatine concentration
of a sample.
If one were using a kinetic rate method, this
methodology, in effect, allows one to "calibrate
out" any temperature effects on the enzymatic
reaction rates. In the case of an endpoint or
equilibrium measurement, effects of temperature in
the Nernst equation could be normalized.
$he porous rods 17 between the electrodes can
be conductive or nonconductive; the use of
1257331
-24-
conductive or nonconductive material between a pair
of electrodes is determined by the analysis to be
carried out.
The porous material can be of a microporous
5 plastic such as nylon, polypropylene, polyethylene,
polyvinylidene fluoride, ethylene-vinyl acetate,
styrene-acrylonitrile or polytetrafluoroethylene.
The porous material can also be cellulosic in ~ature
such as rolled filter paper or longitudinal fiber
bundles. In one embodiment the porous material
consists of crylindrically-shaped pieces of ultra-
high molecular weight microporous polyethylene
(POREX Technologies, Fairburn, GA). This material
is by nature hydrophobic such that it will not wet
15 with water.
The porous material can be rendered hydrophilic
by soaking it in a solution (about 0.1% to 0.6~ by
volume) of a surfactant. Surfactants which have
been found to be useful are the nonionic surfactants
such as the octylphenoxypolyethoxv~hanol family of
surfactants which includes TRITON X-100* (Rohm & Haas
Co.), polyoxyethylene ethers such as BRIJ 35 k
polyoxyethylene sorbitan derivatives such as TWEEN
20* fluoraliphatic polymeric esters such as 3M's
25 FLUORAD FC-171* and surfactants such as Sherex
Chemical Co.'s AROS~ 66 PE-12*, ~nionic and
cationic surfactants can also ~e used. The porous
material without wetting agent or surfactant is
nonconductive. The absence of porous material
between two electrodes results, of course, in
nonconductivity.
* Trademark
~2
-25-
The ion-selective electrodes are comprised of
an ion-selective membrane 18 and a reference elec-
trode 22 which is a silver/silver chloride material.
The ion-selective membrane 18 is made of an ion-
selective compound (an ionophore), a thermoplasticresin and a plasticizer, all of which are soluble in
an organic solvent. It is held in place in the
ion-selective electrode by a retainer means, such as
retainer ring 24. There is, thus, a mechanical seal
between the retainer means and the lower section of
the sensor body such that there can be no leakage
from the membrane. An important characteristic of
the sensor is that the ion-selective membranes
incorporated in it do not require preconditioning
(e.g., soaking in a solution of the ion to be
measured) before the sensor can be used.
In the case of the hydrogen ion-selective
membrane, the ion-selective compound which is incor-
porated into the membrane has the general formula:
R
R~ =C~
I R,
In this compound, Rl, R2 and R3 represent components
of the compound which can be the same or different.
Rl, R2 and R3 can be: 1) a halogen; 2) an alkyl
group having 4-18 carbon atoms; 3) a halogen-sub-
stituted alkyl group; 4) an alkoxy group; 5) ahalogen-substituted alkoxy group; 6) an acid group
~257;~
-26-
represented by -CO2R4 wherein R4 is an alkyl having
- 1-18 carbon atoms; or 7) a keto group represented by
-COR5 wherein R5 is selected from the groups defined
for R4; or 8) a hydrogen atom.
Derivatives of this general formula which are
particularly preferred as components of the
hydrogen-ion-sensing membrane include:
a. 2-trifluoromethyl-4-octadecyloxyphenyl-
hydrazone mesoxalonitrile
18 H370 ~ 1 = C~
~Cf3 `C~
b. 2-octadecyloxy-5-trifluoromethylphenylhydrazone
mesoxalonitrile
¦ C~3 ~ _ ~ H ~~c~
~ Cl8H3 7
c. 2-octadecyloxy-5-carbethoxyphenylhydrazone
mesoxalonitrile; and
C~ ~50--c~
N H` ~1- C
~O `C~
Cl~ ~ 7
d. 2-octadecyloxy-4-fluorophenylhydrazone
mesoxalonitrile.
F~)~ - c~
C/8 ~37
~25733~L
-27-
Membranes in which the ion-selective compound
- is carbethoxyphenylhydrazone mesoxalonitrile are
particularly useful in that they exhibit response
times of one minute or less; yield stable potentials
and exhibit slopes on the order of 57-59 mv per
decade change in hydrogen ion concentration and do
not require any preconditioning. It is not clear
why such membranes have excellent performance
characteristics. However, it might be attributable
to the presence of an electron-withdrawing ester
group meta to the anilino group and its proton,
which is apparently more acidic and, therefore, more
easily exchanged.
In addition to the ion-selective compound
described above, the ion-selective membrane 18 of
this invention also comprises a plastic or a thermo-
plastic resin and, optionally, a plasticizer. The
plastic used can be any plastic which can be used to
form a film by solvent casting. For example, the
plastic can be polyvinyl chloride, polyvinyl ace-
tate, silicone rubber or cellulose acetate. This
membrane component serves the purpose of providing
support and form to the membrane, and acts as a
matrix into which the ion-selective compound is
incorporated. In addition, it serves as a barrier
to water because it is a hydrophobic material. In
one embodiment, the thermoplastic resin used in the
membrane is polyvinyl chloride. Others which can be
used include: cellulose acetate, polyvinyl acetate
and silicone rubber.
i257~
-28-
~he plasticizer, which is an optional component
- of the membrane, can be any nonvolatile material
suitable for the general purpose of facilitating the
compounding or production of the membrane formula-
tion and improving the membrane's flexibility. It
also contributes to the dissolution of the iono-
phore. It can be, for example, one or more of the
following, used alone or in combination: phthal-
ates; adipates; sebacates; aliphatic and aromatic
ethers; aliphatic and aromatic phosphates; aliphatic
and aromatic esters; and nitrated aliphatic and
aromatic ethers. In one embodiment of this inven-
tion, the plasticizer is 2-nitrophenyloctylether.
The components of the ion-selective membrane 18
can be present in varying amounts. The plastic used
can comprise about 10-30% by weight of the membrane
and in one embodiment is about 28% by weight. The
ion-selective compound can be from about 1% to about
10% by weight of the membrane and in general will be
about 3% to about 6% by weight. The plasticizer can
be from about 50% to about 80% by weight of the
membrane and generally comprises about 70% by
weight.
The ion-specific membrane 18 incorporated into
the ion-selective electrodes of this invention will
generally be of a thickness greater than 1 mil and
preferably will be from about 3 to about 15 mils in
thickness.
The ion-specific membrane of this invention can
be made by at least two different methods: solvent
casting and dip coating. Solvent casting is
~2.~7~31
-29-
illustrated in Example 2 and in general includes the
following steps:
1. preparation of the hydrogen ion-specific
compound by: a) alkylation o~ a nitrated phenolic
material to produce an ether; b) reduction of the
nitro group to an amino group; c) precipitation of
the hydrochloride salt of the amine; and d) con-
version of the hydrochloride salt to the
mesoxalonitrile derivative;
2. dissolution of the ion-specific compound; a
plasticizer and a plastic material in a volatile
(organic) solvent; and
3. removal of the volatile organic solvent
(e.g., by evaporation).
After the membrane has been formed, it is
shaped (e.g., by cutting, use of a punch-type
instrument, etc.) into an appropriate configuration
and incorporated into an electrode body, such as the
eiectrode ~ody of the present invention or an ORION*
electrode body. (Orion Research, Inc., Cambridge,
MA. catalogue #950015). Figure 5 shows the ORIO~-
electrode body 61 with the ion-selective membrane 62
of this invention, held in the electrode by means of
an O-ring 63. A second O-ring (not shown) prevents
sample leakage around the membrane.
The ion-selective membrane can also be formed
by means of a dip coating or a continuous coating
process method. In this method, a base material,
such as a nylon mesh or a polyester mesh (e.g.,
PECAP 355*, Tetko, Inc., Elmsford, NY)) is placed in
`~ a solvent bath (e.g., ketone, such as 4 methyl-2
; * Trademark
~; "s,.
....
~257331
-30-
pentanone) and the air trapped within the mesh
- removed. This can be done, for example, ultrasoni-
cally using a sranson ultrasonic cleaner; treatment
takes about 10-15 seconds. While still wet with
solvent, the mesh is placed in the membrane formula-
tion; after it is removed, it is scraped to remove
excess formulation. The mesh containing the formu-
lation is then dried; this can be done, for example,
at room temperature or in warm air (e.g., 35-50 C).
The internal reference material consists of a
solution of the salt of the ion to be determined.
The counter ion consists of chloride if the refer-
ence electrode is silver/silver chloride. A gelling
agent such as agar or agarose can be used to fix the
reference material in place. In the case of the pH
sensor, the reference material is a buffer solution.
In one embodiment, the pH reference material con-
sists of 50 ml glycerol, 50 ml of an aqueous solu-
tion consisting of pH 7.4 phosphate-buffered saline
(SIGMA Chemical Co., St. Louis, MO) and 2.0g of
agar. The mixture is heated to dissolve the agar
and then placed in the well behind the membrane.
After a period of time, normally 5-10 minutes, the
mixture gels. For a potassium sensor the aqueous
solution can be 50 ml of 0.2 M KCl; for the sodium
and chloride sensors, the aqueous solution can be
0.2 M NaCl.
To modify the sensor of Figures 1-3 for the
determination of other ions, it is necessary to
change at least one of the membrane components and
to modify the internal reference material to include
1257331
-31-
the ion to be determined. For example, for potas-
sium determinations, the membrane would include an
ionophore specific for potassium, such as valinomy-
cin, a crown ether derivative of benzo-15-crown-5 or
some other neutral carrier. The internal reference
material would consist of a potassium chloride
solution with or without a gelling agent.
To modify the sensor of Figures 1-3 for the
determination of sodium, the membrane would include
an ionophore for sodium, such as monensin, a crown
ether derivative of benzo-12-crown-4 or some other
sodium neutral carrier (e.g., as described by Simon
et al., Analytical Chem., 51: 351-353 (1979)). The
internal reference material would include sodium
chloride.
To modify the sensor of Figures 1-3 for the
determination of chloride, the membrane would
include an ionophore for chloride such as the
quaternary ammonium salt Aliquat 336. The internal
reference material would include chloride ion.
As mentioned previously, sensor 30 has a handle
5 which has a magnetic stripe 8. This stripe is
very similar to the stripes commonly found on credit
cards and bears data recorded onto the stripe at the
time of manufacture. The data includes, for ex-
ample, the following information:
;~ 1. Test type: A unique two digit number
identifies each specific type of test card
(i.e., each specific analyte). This
eliminates the need for operator entry of
,
~ .,.
~257~3~
-32-
test type information and consequently
eliminates possibilities of error.
2. Calibration constant: Three digit numeri-
cal constant used in an equation stored in
the instrument memory to calculate analyte
concentration based on the measured
signal. This constant may be used in
conjunction with the simultaneous calibra-
tion to determine possible test card
failure.
3. Expiration date: Four digit code that
enables the instrument to determine the
expiration date of the test card and
ensure that no out-of-date cards are used.
The format for this constant is YYWW where
YY is the last two digits of the year
(e.g., 85 for 1985), and WW is the week of
the year (e.g., the week of May 5 through
May 11 is week 19).
4. Reserved: Four digits are reserved for
future use.
In addition to the data described above the follow-
ing "housekeeping" data can be written on the card:
1. Leader: Minimum of 15 "O"s for synchroni-
zation of the instrument magnetic stripe
reader circuitry.
2. Start sentinel: Special character used to
signify the start of data.
3. LRC: Longitudinal Redundancy Check used
in conjunction with a parity bit attached
to each data character to detect possible
~25733~
-33-
errors in reading data from the magnetic
- stripe.
An independent function of the magnetic stripe
is to prevent reuse of the disposable test cards
which are designed for single use only. This is
accomplished by destroying the coded information as
the test card is removed from the instrument. The
error checking functions described above will reject
the card if an attempt is made to reuse it.
Figure 6 shows a sensor of the present inven-
tion which can be used to determine the activity or
concentration of components of samples by means of a
differential measurement technique. Components of
biological fluids (e.g., blood, serum, plasma,
urine, saliva, cerebrospinal fluid) which can be
measured include, for example, glucose, urea,
triglycerides, creatinine, lipase, uric acid and
other enzymes (e.g., alanine aminotransferase (SGPT
or ALT); alkaline phosphatase (ALP); creatine kinase
~CK) and aspartate aminotransferase (SGOT or AST)).
The sensor 70 is comprised of ion-selective
electrodes, which are held in a frame or body 40
which has an upper section 42 and a lower section
44. Upper section 42 of frame 40 has four openings
34 and three grooves 36, which are positioned
between the openings. Lower section 44 of frame 40
has four openings 35 and three grooves positioned
between the openings. Upper section 42 and lower
section 44 are in such a relationship that openings
34 and openings 35 are aligned and grooves 36 of the
upper section 42 and the grooves in the lower
-34-
section 44 are aligned. As a result, openings 34
- and openings 35 form openings 39 and grooves 36 of
the upper section 42 and the grooves in the lower
section 44 define spaces between three of the
openings 39. Upper section 42 has small holes 43
located between openings 34. Holes 43 allow air to
enter space 41.
The ion-selective electrodes are located within
the openings 39 and are connected by a porous
material 46 which is cylindrical or rod shaped.
Porous material 46 provides a means for ionic flow
between the electrodes upon the application of a
sample at the electrodes. The cylindrical or
rod-shaped porous material 46 is located in the
spaces 41 between openings 39.
As described for the sensor of Figures 1-3,
fewer than or more than the four ion-selective
electrodes depicted in Figure 6 can be used. Also
as previously described, the porous material between
the electrodes can be hydrophilic or hydrophobic;
the type of analysis determines this characteristic.
For example, if an enzyme substrate is to be mea-
sured using the sensor represented in Figure 6 and a
reference sample is to be used, then the configura-
tion could be as shown in Figure 4d.
The ion-selective electrodes are comprised of a
layer 48 having enzyme or substrate immobilized on
it; an ion-selective membrane 50; a membrane 52
placed between the two; an internal reference
material 54 and a reference electrode 56. There can
~2~;73:~1
-35-
be an additional membrane 58 in proximity with layer
- 48.
The ion-selective membrane can be selective for
H or ammonium ions (NH4 ), as in the case of a urea
electrode. In the case of the ammonium-selective
membrane, the ionophore is an ammonium-selective
material such as nonactin. The ion-selective
membrane is held in place in the ion-selective
electrode by a retainer means, such as retainer ring
60.
Layer 48 has at least one enzyme or substrate
immobilized on it. Layer 48 can be, for example, a
nitrocellulose membrane (e.g., type AE100 (trademark),
Schleicher and Schuell, Inc. Keene, NH) having a
thickness of about 180 microns. It can also be made
of nylon, cellulose, cellulose acetate, glass fiber
or any porous material which can serve as a solid
support for the immobilization of the enzyme or
substrate. The enzyme or substrate can be immobil-
ized by physical restraint (e.g., adsorption ontothe solid phase) or by chemical restraint (e.g.,
covalent attachment or crosslinking).
Layer 48 is separated from the ion-selective
membrane 50 by means of a membrane 52, which is
generally thin (e.g., less than 20 microns thick)
and can be, for example, a dialysis membrane te.g.,
Spectra Por 2, Spectrum Medical Industries, Inc.,
Los Angeles). Membrane 52 serves the purposes of,
for example, preserving the activity of the immobil-
ized enzyme of layer 48 by preventing the migrationof components of the ion-selective membrane 50 into
~2S73;~1
-36-
layer 48; migration might cause denaturation or
- inactivation of the enzyme.
Membrane 58 is optional and its presence will
be determined by what component of a sample is being
measured and/or how that component is being mea-
sured. The function of membrane 58 is to aid in the
physical restraint (immobilization) of the enzyme or
substrate of layer 48 and/or to serve as a diffu-
sional barrier to the substrate. For example, to
measure the concentration of a substrate of the
enzyme immobilized in layer 48, it may be desirable
to make a kinetic measurement, rather than an end
point measurement. To extend the range of linearity
for a kinetic measurement, it is necessary that the
enzyme reaction be diffusion limited. In this
instance membrane 58 will serve both purposes. If,
however, an end point measurement is desired a
diffusional barrier would not be required and
membrane 58 is not needed. Membrane 58 can be made
of regenerated cellulose (dialysis membrane) or
other ultrafiltration or reverse osmosis membranous
material which can serve as a diffusional barrier to
the substrate.
The internal reference material 54 is comprised
of a buffer solution. In one embodiment, it is
comprised of a filling solution of pH 7.4 phosphate
buffered saline. In a preferred embodiment, it is a
gel comprised of about 50 ml. pH 7.4 phosphate
buffered saline, about 50 ml. glycerol and about 2.0
g agarose.
~257331
-37-
Reference electrode 56 is an electrode with a
fixed potential which exhibits no variation in
liquid junction potential when the sample solution
is varied or replaced by a calibrating solution.
The main requirements for satisfactory reference
electrode performance are reversibility (in the
electrochemical sense), reproducibility and stabil-
ity. Three types of reference electrode systems are
in current use: metal amalgams, such as the
saturated calomel electrode (SEC); redox couples,
such as a quinhydrone or ferri-ferrocyanide elec-
trode; and metal/metal halide electrodes, such as
the silver-silver chloride reference electrode. In
the present invention, the reference electrode will
generally be a silver/silver chloride button.
The subject of this invention is illustrated by
the following examples, which are not to be consi-
dered limiting in any way.
Example 1 - Preparation of the Hydrogen Ion SPecific
ComPound 2-octadecYloxy-5-carbethoxyphenylhydrazone
mesoxalonitrile
The hydrogen ion specific compound 2-octadecyl-
oxy-5-carbethoxyphenylhydrazone mesoxalonitrile is
prepared in the following manner. First, ethyl-4-
octadecyloxy-3-nitrobenzoate is prepared according
to the general procedure described in Organic
Synthesis, Coll. vol. 3, pp 140-141.
The following mixture is refluxed for 72 hours
in a l-liter-3-neck flask fitted with a mechanical
stirrer:
~ - ~
~25733~
-38-
25g (0.12 moles) of ethyl-4-hydroxy-3-
- nitrobenzoate
Lancaster Synthesis, Windham, NH
catalogue #6451
41g (0.12 moles) of 96% l-bromooctadecane
Aldrich Chemical Co., Milwaukee, WI.
catalogue #19,949-4
16.4g of anhydrous potassium carbonate
500 ml. of dry acetone
At the end of the reflux period, the reaction
mixture is cooled to room temperature (i.e., about
20-27C) and filtered to remove salts. The filtrate
is evaporated using a suchi-Brinkman rotary evapora-
tor. The residue is dissolved in 500 ml of toluene
and washed in turn four (4) times with 400 ml of 5%
aqueous sodium hydrogen carbonate and then one (l)
time with 400 ml of saturated aqueous sodium chlor-
ide. The toluene layer is dried over anhydrous
sodium sulfate and evaporated to dryness; this
produces an oil which crystallizes on standing. The
solid is recrystallized from hexane with charcoal
clarification; the yield is 36g (66% yield) of
ethyl-4-octadecyloxy-3-nitrobenzoate (m.p. 48-56).
The nitrocompound is reduced to the amino
derivative by catalytic hydrogenation. To 200 ml of
hexane is added lOg of the above nitro compound
along with 0.5g 10% palladium on carbon. The
mixture is placed on a Parr hydrogenator, heated to
40C under 60 psig of hydrogen and shaken in the
machine for 18 hours. The reaction mixture is
filtered through Celite filter aid medium (usually
~2573:~1
-39-
diatomaceous earth (Manville)) while hot (e.g.,
about 50C), and the solution is treated with HCl
gas to precipitate the hydrochloride salt of the
amine in nearly quantitative yield (m.p. 141-145).
The hydrochloxide salt of ethyl-4-octadecyloxy-
3-aminobenzoate is converted to the mesoxalonitrile
derivative according to a modification of the
procedure of Brown et al. (U.S. Patent No.
3,743,588). 8.5g (0.018 moles) of the above hydro-
lO chloride salt is dissolved in 1 liter of dimethyl
formamide (DMF) with warming and brought to a
temperature of about 0. 3 ml of concentrated HCl
is added to the cold solution, followed by 1.3g of
sodium nitrite dissolved in 500 ml of DMF. The
15 reaction mixture is stirred magnetically for 1 hour.
1.15 ml of prewarmed malononitrile is then added and
the mixture is stirred magnetically for about 10
minutes. Triethylamine is added in sufficient
quantity to produce a strongly basic solution (e.g.,
20 pH greater than 9). The mixture is allowed to warm
to room temperature and stirred for 18 hrs. The
reaction mixture is acidified (to a pH of about 2.3)
with concentrated HCl, and a precipitate results.
The precipitate is filtered, washed with water and
25 recrystallized from methanol. The result is about
7.0g, (76%) of 2-octadecyloxy-5-carbethoxyphenyl-
hydrazone mesoxalonitrile tm.p. 76-77). The other
derivatives shown in Table 1 were prepared in a
similar manner.
~ ~z57~3~
--40--
o~
~ _ o ~ a~
r I ~ u~
V) C~
_ ~
t~ a~
a~
~ V N O~
_ 3 0
.
~e--
L C~
L~JO~_ U~ ~ ~ ~
JC~ 3 ~ C~
~ O
S C
~ O
~ _, 1.~, . ~
~ . ~
O ,.
N ¦ O
O
~` 1-- r~
Co ~ I
_ N ')
E .
12~;733~
-41-
Example 2 - Solvent Casting of a PVC Membrane and
- Incorporation Into an Ion-Selective Electrode
1.5 ml of 2-nitrophenyloctylether (Fluka
Chemical Corp., Hauppauge, NY, cat #73732) and 0.10g
of the hydrogen ionophore (such as that prepared
according to the method illustrated in Example 1)
are dissolved in 10 ml of tetrahydrofuran. To this
solution is added 0.6g of powdered polyvinyl chlor-
ide of very high molecular weight and of density of
1.385 g/cc (Aldrich Chemical Co., Milwaukee, WI, cat
#18,261-3). The mixture is shaken (e.g., by vortex-
ing) until the PVC is dissolved. The PVC solution
is poured onto a 1/4" thick sheet of stress-relieved
polypropylene and allowed to remain under conditions
15 which result in the evaporation of the tetrahydro-
furan. For example, evaporation can occur at room
temperature under a fume hood so that the tetra-
hydrofuran is removed as it evaporates. The product
formed is a plastic membrane in which the ion-
20 specific compound is incorporated. Disks are cutfrom the membrane using a #7 cork borer (id 0.5in)
and mounted in an Orion electrode body (Orion
; Research Inc., Cambridge, MA, cat #950015). The
internal reference electrode is a silver/silver
25 chloride reference electrode and the internal
filling solution is pH 7.4 phosphate buffered saline
~SIGMA Chemical Co., St. Louis, MO, cat #1000-3).
Assessment of the pH-sensing capability of the
electrodes was carried out. All measurements were
30 made versus a double junction Ag/AgCl reference
electrode (Corning, cat #476067) in 250 ml beakers
.~
~25733~
-42-
with magnetic stirring. For comparison a combina-
tion pH/reference electrode was dipped into the test
solution to record pH changes, such that mv changes
in the membranes can be related to pH changes
measured with the glass electrode. The membrane of
the present invention exhibited good mechanical
strength, as indicated by the fact that it did not
tear or break when stretched. It also exhibited
good analytical performance, yielding slope values
of 55-58 mv per decade over the pH range of 6 to 8
with performance decreasing somewhat on either side
of this range. For a monovalent ion, the theoreti-
cal slope is +59.1 mv at 25C. In addition, the
response time or the time to reach equilibrium was
less than one minute and the electrodes required no
preconditioning.
The properties of the above membrane were
compared to those of a membrane prepared in an
identical manner except that the hydrogen ion
selective component was the previously described
3-chloro-4-octadecyloxyphenylhydrazone mesoxaloni-
trile. In contrast to the previous membrane, the
membrane prepared from the chlorinated mesoxaloni-
trile gave diminished performance; slope values were
40 mv or less over the same pH range, indicating a
loss in sensitivity.
The performance of the various hydrogen iono-
phores is summarized in Table 1. The results show
that the three derivatives with the 18 carbon
lipophilic chain ortho to the ionophoric site
outperformed the derivative in which the lipophilic
12S7~3~
-43-
chain was para. This suggests that perhaps the
lipophilic chain adjacent to the site of hydrogen
ion exchange enhances the response towards hydrogen
ions. The lipophilic chain could form a lipophilic
pocket around the active site, thereby increasing
the selectivity towards hydrogen ions.
Example 3 - Preparation of an Enzyme Electrode for
the Measurement of Urea
A sensor as shown in Figure 6 is used for the
determination of urea. A urea-selective sensor is
made by incorporating a second membrane containing
the enzyme urease in the ion-selective electrodes of
the sensor, which can be either a pH or an ammonium
ion-(NH 4) membrane. Urease-containing membranes
are present in electrodes a, b and c shown in Figure
4f. The two membranes are separated by a third
membrane (e.g., membrane 52 of Figure 6~.
The sample to be tested (e.g., blood, plasma),
a reference and a calibrant sample are added at the
electrodes as shown in Figure 4e.
A and B are wicks made of porous material and
nave been made conductive by the addition of a
wetting agent. a, b and c are ammonium ion-selec-
tive electrodes. Two different known levels of
urea, reference and calibrant are added at a and b,
which have urease. Patient sample is added at c,
which also has urease.
~25~
-44-
Reactions catalyzed by urease occur at b and c.
- This allows the comparison of the patient sample
with two calibrant samples, thus allowing for the
calibration of the test card.
The following reactions occur where urease is
present:
NH2CONH2 ~ H2O ~2NH3 f C2
NH3 ~ H2O~ NH4 ~ OH
C2 ~ H20~ HC03 H
10 As a result, an ammonium selective membrane can be
used to detect the change in ammonium ion (NH4 )
concentration or a pH selective membrane can be used
to detect the increase in pH which occurs because of
the production of OH .
A urease containing membrane is prepared as
follows: A 5 mg/ml urease (300 units/mg New England
Enzyme Center, Boston, MA) solution is prepared in
pH 7.4 phosphate buffered saline (Sigma Chemical
Co., Cat. # 1000-3): other buffers found to be
20 useful are 10 mmol/L sodium Hepes in 0.1 M MaCl (pH
7.4) and 10 mmol/L NaH2PO4 10 mmol/L tris base in
0.1 M NaCl (Ph 7.4). A 47 mm diameter disk of 12
micron pore size nitrocellulose membrane (Schleicher
& Schuell, Inc., Keene, NH, 03421 Grade AE 100) is
soaked in 700 ul of the enzyme solution for 5
minutes at room temperature. The membrane is
removed from the solution and the excess fluid
scrapped off with a stirring rod. The membrane is
placed on a hydrophobic surface (e.g., a sheet of
30 polypropylene) and air dried at room temperature for
30 minutes. The membrane is stored in a closed
container at 4C. Disks are cut from the membrane
l~S73;~1
-45-
and placed in the electrodes at a, b and c; see
- Figure 4f. Sample solution (25 micro liters) is
placed in c calibrating solutions can be placed in a
and b. The urease in a, b and c acts on the urea in
the calibrants and sample, liberating both ammonia
and carbon dio~ide (CO2). The increase in pH or
ammonia is measured.
For example, a pH 7.5 buffer was prepared
containing 40 mmol/L NaH2PO4, 40 mmol/L tris base
10 and 100 mmoltL NaCl. Buffer solutions containing 5
mmol/L urea and 10 mmol/L urea were prepared. After
approximately one minute a 25 mv difference was
observed when the 5 mmol/L urea solution was placed
in a and c and buffer containing no urea was placed
in b. A 50 mv difference was observed with the 10
mmol/L solution of urea. No difference in emf was
observed in the absence of urea.
Similar behavior was observed when the pH
membrane was replaced with an ammonium selective
20 nonactin membrane.
Example 4
Preparation of an Enzyme Electrode for the
Measurement of Glucose
Referring to Figure 4e, cells a, b and c
each contain a membrane with immobilized glucose
oxidase and catalase.
The following reactions occur where glucose
oxidase (GOD) and catalase are present:
Glucose ~ H2O~O2 GOD ~ Gluconic acid (H ) ~ H2O2
H202 catalase-~?l/2 2 ~ H20
-
lZ57331
-46-
The purpose of the catalase is to recycle the oxygen
-consumed in the oxidation of glucose, thereby
extending the linear range within which glucose may
be measured. The glucose oxidase/catalase membrane
was Prepared as follows: The previously described
nitro cellulose membrane was soaked in 700 micro
liters of a solution consisting of S mg/ml glucose
oxidase (300 units/mg, Boehringer Mannhein) and 1
mg/ml catalase (40,000 units/mg Sigma Cat. # C-100)
in pH 7.4 phosphate buffered saline. The membrane
was dried as in example 3. Enzyme disks were placed
in a, b and c. The concentration of glucose in the
sample was found to be proportional to the potential
diference developed between b and c.
15Another method for the determination of glucose
uses the hexokinase/ATP as the enzyme system.
Referring to Figure 4e, cells 1, b and c each
contain a membrane with immobilized hexokinase, ATP
and a magnesium salt such as the chloride, acetate
20 or sulfate.
The following reactions occur where hexokinase,
ATP and magnesium ion are present:
Glucose ~ ATP Mg > Glucose-6-Phosphate ~ ADP ~ IT
~ The hexokinase/ATP membrane was prepared as follows:
;25 The previously described nitrocellulose membrane was
soaked in l.Oml of a solution consisting of lOg/L
;magnesium acetate, 50mg/ml ATP, and lOmg/ml hexo-
kinase in a buffer consisting of 10% glycerol, 0.25%
Triton X-100 and 0.005M triethalnolamine pH 7.8.
30 The concentration of glucose in the sampel was found
125q~3~
-47-
to be proportional to the potential difference
- developed between b and c.
Example 5
Preparation of an Enzyme Electrode for the
Measurement of Triglycerides
Referring to Figure 4j, membranes contain-
ing enzyme, in this case lipase, are placed in cells
a, b and c. The following reaction occurs where
lipase is present:
10 Triglycerides ~ H2O lipase ~ glycerol ~ fatty acids (H )
As in the previous examples the membranes were
prepared by soaking a nitrocellulose membrane in a 5
mg/ml. buffered solution of lipase (Sigma Chemical
Co., Cat. # L4384). The decrease in pH which occurs
15 upon addition of a sample containing triglycerides
is detected with the pH electrode.
Example 6
Preparation of an Enzvme Electrode for the
Measurement of Uric Acid
As in the previous examples and referring to
Figure 4e, an electrode specific for uric acid is
prepared by soaking a nitrocellulose membrane in a
solution of uricase (50 units/ml Sigma Chemical Co.,
Cat. # U1878) plus catalase (1 mg/ml 40,000 units/mg
Sigma Chemical Co., Cat. # C-100). The enzyme
membrane is placed in cells a, b and c.
The following reactions occur where uricase and
catalase are present:
-
~2573~1
-48-
uric acid ~ 2 ~ H2O uricase Allantoin ~ CO2 ~ H2O2
- H22 catalase ~ 1/2 2 t H2O
The CO2 produced in the reaction causes a decrease
in pH which is detected with the pH electrode.
Example 7
Preparation of an Immunosensor for the
Measurement of Theophylline
In the present case, the measurement of
theophylline is based on differences in
immunochemical reactivity between theophylline and
caffeine. Caffeine differs from theophylline by a
methyl group. Theophylline and caffeine have been
coupled to a crown ether moiety and each conjugate
has been incorporated into a PVC membrane. The
theophylline membrane forms one half-cell while the
caffeine membrane forms the other half-cell. The
caffeine membrane serves as a reference electrode
and separately, each membrane exhibits a response to
potassium ion. However, when each membrane is
incorporated into a concentration cell and the same
solution containing potassium ion is placed in each
half-cell, the emf developed between the two half-
cells is very close to zero millivolts. Although
the theophylline and caffeine conjugates exhibit
identical behavior in the electrochemical sense,
they differ greatly in their immunochemical re-
activity towards theophylline antibody. This
difference in immunochemical reactivity and identi-
cal ionophoric electrochemical reactivity, form the
basis for a competitive binding differential
potentiometric assay for theophylline. This method
~2S7331
-49-
does not require constant ionic strength nor con-
- stant potassium ion activity ~as in the Rechnitz
approach). A major advantage of the present method
resides in the fact that undiluted and undialyzed
serum samples may be used. The arrangement of the
sensor can be as represented in Figure 41.
The sensor consists of theophylline crown
ether conjugate membranes in cells a and d and
caffeine crown ether conjugate membranes in cells b
and c, separated by porous junctions A and C. The
sample containing theophylline is added to c and d
while a calibrant is added to a and b. The competi-
tive binding aspects require the presence of theo-
phylline antibody with a low cross reactivity
towards caffeine. The antibody may be added to the
sample and calibrant before analysis in the sensor
or the antibody may be contained within the sensor
cells such that competitive binding occurs within
the sensor. The competition for antibody site is
, 20 set up between theophylline immobilized in the
membrane and theophylline in the sample:
Ab ~ Ag(sample) ~ Ag-l~~~~AbAg(sample) ~ AbAg-1
(membrane) (membrane)
As theophylline antibody binds to the theophylline
immobilized in the membranes (a and c) a change in
potential develops, the magnitude of which is
inversely proportional to the concentration of
theophylline in the sample. So as the theophylline
concentration in the sample increases, less antibody
is available to bind to the membrane, and therefore,
~Z~;733~
-50-
less of an emf is developed. For a typical assay,
- potential changes due to antibody binding are
normally on the order of a few millivolts. Another
advantage in using the differential approach with a
theophylline membrane, versus or in combination with
a caffeine "reference" membrane, is that the common
mode or initial starting potential is near zero
millivolts and the signal to be measured is on the
order of millivolts. The low common mode potential
allows for greater senstivity in the measurement of
the signal voltage and therefore a more accurate and
precise signal measurement is possible when the
background or com~on mode voltage is less than the
signal voltage. As previously discussed, the
potential difference between the theophylline and
caffeine half-cells is nearly zero when treated with
the same potassium ion-containing sample. However,
when either half-cell is replaced with a reference
electrode such as an SCE or a silver-silver chloride
reference electrode, the potential difference with
the same potassium ion solution increases to as much
as 50 mv or more.
The synthesis of the conjugates and the method
of preparation of the membranes is described below:
~25q331
Reaction Scheme for the Preparation of Theophylline
Benzo-15-Crown-5-Conju~ate
H3C~ NH O ICH3 H3C~ ~ o ~ H3
1 ) o"~J --CICH2O--C--C--cH3 DMF > ~ N CH3
I -mahyl-7-1(pivaloylo~y)mclhyl]~ltnlhine
I -melhylbl-nlh;n~
2) 1 ~ 3 + HNO3 HOAc ~
~/ ~/ 4'-nilrobenzo 15-crown 5
Benzo-lS-crov~n 5 (R-ney nickcl~
I H2/Ni DiCI amine
\ 2 HCI (g)
c~ N~[~
4'~mino benzo-15-crown~5 hydrochloride
~ NH3CI- + Br(CHI)l--C--OH + BoPCI
3) o ~ n ~ 3-20 1l
~-- J 11 1 u
O /~ H 1l
~,N--C--(CH2)"Br
~-J
4' Benzo.15 crown-5
( I l-bromoundect noamide)
~5~3~
-continued
Reaction Scheme for the Preparation of Theophylline
Benzo-15-Crown-5-Conjugate
</~ /~ HN_C_(CH2)~Br + H3~ ~ 1l Cl H3
O ~ H
H3C~ O CH3
O~ N N--CH2--O--C--C--CH~
tCH2)~ 1~\ o ~\
tH~ O >
~o
HlC~ O ICHl I ~ OH HlC~
~ N~ N--CH2--O--c--c--cH3 2 H o > N ~j`lH
(1H"~ I~\ O ~\ (CH2)~ ~\ O ~\
C--N~ > ~ O
~o ~ ~
Theophylline B~nzo-crown 5 Conju8~1e
-51a-
~2S7331
-52-
The first step in the synthesis involves
blocking the 7 position in l-methyl xanthine. The
procedure of Hu, Singh and Ullman, J. Am. Chem.
Soc., 45: 1711-1713 (1980) was followed. To 700 ml
of dry DMF in a l-liter conical flask was added 3 2g
(18.9 mmoles) of 98~ l-methylxanthine (Aldrich
Chemical Co., cat. # 28,098-4), the mixture was
stirred magnetically and warmed until dissolution
occurred and then cooled to room temperature. To
the stirred solution was added 2.0g (18.9 mmoles) of
anhydrous sodium carbonate followed by the dropwise
addition, over a period of 1 hour, of 3.0g (19.3
mmoles) of 97% chloromethyl pivalate (Aldrich
Chemical Co., cat. #14,118-6) dissolved in 50 ml of
DMF. The reaction mixture was stirred overnight at
room temperature. The reaction mixture was evapor-
ated to dryness and the residue treated with 50 ml
of methylene chloride. The insoluble material was
collected by filtration and washed with lN HCl to
give 1.24g of l-methylxanthine. The methylene
chloride solution was evaporated to dryness to yield
a mixture of mono- and bis- protected xanthine. The
mono-protected l-methyl-7-,'(pivaloyloxy)methyl~
xanthine was separated from the bis protected
xanthine by crystallization from ethyl acetate. By
this method there was obtained 1.5g of the mono-
protected xanthine (m.p. 204-206). The ethyl-
acetate filtrate from the isolation of the mono-
protected xanthine, containing the bis-protected
xanthine, was evaporated to dryness and refluxed
with 20 ml of 2N sodium hydroxide for 18 hvurs. The
~25~33~
-53-
reaction mixture was cooled and acidified with
concentrated ~Cl and extracted with chloroform.
Evaporation of the aqueous layer yielded 0.5g of
l-methylxanthine.
The second step in the synthesis is the prepar-
ation of 4'-amino-benzo-15-crown-5 hydrochloride.
To a stirred solution of 30g (6.11 mmoles) of
benzo-15-crown-5 (Parish Chemical Co., Orem, UT,
cat. #I405) in 800 ml of l:l glacial acetic acid/
chloroform was added dropwise, over a period of 1
hour, a solution of 15 ml concentrated nitric acid
dissolved in 50 ml of glacial acetic acid. The
reaction mixture was stirred for one hour and then
evaporated to dryness on a rotary evaporator. The
residue was treated with 500 ml of 5% aqueous sodium
hydrogen carbonate and the mixture partitioned with
500 ml of chloroform. The chloroform layer was
washed with saturated aqueous sodium chloride and
dried over anhydrous sodium sulfate. Evaporation of
the chloroform yielded 30g (86%) of 4'-nitobenzo-15-
crown-5. The product was recrystallized from
ethanol and 15g (48 mmoles) dissolved in 250 ml of
dioxane and catalytically reduced over Raney nickel
(Aldrich Chem. Co., cat. #22,167-8) in a Paar
hydrogenator at 50C.
The reaction mixture was filtered and the
filtrate evaporated to dryness and the residue
dissloved in 200 ml of ethyl acetate. The solution
was then treated with HCl gas to precipitate the
hydrochloride salt of 4'-amino-benzo-15-crown-5.
Yield 9.0g (58%).
~ .
:
~25~331
-54-
The third step in the synthesis is the prepara-
tion of 4'-benzo-15-crown-5-(11-bromoundecanoamide).
To 10g (37.7 mmoles) of ll-bromoundecanoic acid
(Aldrich Chemical Co., cat. #s8,280.4) in 100 ml of
S CH~C12 in a 250 conical flask was added 5.5 ml of
triethylamine. The mixture was cooled tG 0C and
9.67g of N,~-Bis['~-oxo-3-oxazolidinyl] phosphordi-
amidic chloride (BOPCL, Chemical Dynamics Corp.,
cat. #12-1370-00) was added. The mixture was
stirred for 0.5 hours then 12.06g (37.7 mmoles) of
4'-amino-benzo-15-crown hydrochloride was added
along with 5.5 ml of triethylamine. To the reaction
mixture was added dropwise another 5.5 ml of
triethylamine in 50 ml of CH2C12 over a period of 1
hour. The reaction mixture was allowed to warm to
room temperature for 18 hours. Water (50 ml) was
added and the mixture acidified with concentrated
HCl. The CH2C12 layer was washed in turn 3 times
with 200 ml saturated NaCl solution and 3 times with
200 ml 5% sodium hydrogen carbonate solution. The
CH2C12 layer was dried over anydrous sodium sulfate
and evaporated to dryness. The residue was
crystallized from hexane/ethyl actetate to yield
8.8g ~44~) of 4'-benzo-15-
crown-5-(11-bromoundecanoamide), m.p. 101-103,
analysis calaculated for C25H40BrN06, C:56.60;
H:7.60, Br:15.06; N:2.62. Found: C:56.41; H:7.46;
Br.:14.89; N:2.60.
In the fourth step, the crown ether was coupled
to the protected l-methylxanthine. To 50 ml of DMF
; in a 125 ml conical flask was added 1.0g (3.57
'
1257331
-55-
mmole) of l-methyl-7-[(pivaloyloxy)methy~ xanthine.
The mixture was warmed (e.g., to about 80C with
stirring) until dissolution. The solution was
cooled to room temperature and 0.76g (7.1 mmole) of
anhydrous sodium carbonate was added followed by the
addition of l.90g (3.57 mmole) of the crown ether
amide. The reaction mixture was stirred under
nitrogen for 3 days. The DMF was removed by rotary
evaporation and 100 ml of water was added and the
mixture extracted 3 times with 50 ml of chloroform.
The combined chloroform extracts were washed 1 time
with 100 ml of saturated aqueous sodium chloride and
dried over anhydrous sodium sulfate. Evaporation of
the chloroform yielded a pale brown oil.
The fifth step is the removal of the protecting
; group with aqueous base. To the above oil in a 250
ml round bottom flask was added 75 ml of 0.2M NaOH
and 25 ml of methyl alcohol. The mixture was
refluxed for 1 hour and cooled to room temperature.
The solution was acidified with concentrated HCl
with cooling in an ice/acetone bath. The precipi-
tated product was recrystallized from ethyl alcohol-
water to yield 1.2g (5S~) of an off-white crystal-
line solid, m.p. 166-170. The material was homo-
genous by hplc analysis. Analysis calculated for
C31H45N508: C:60.47; H:7.37; N:11.37. Found:
C:60.25; H:7.23; N:ll.O9. The benzo-12-crown-4
derivative was prepared by a similar method.
Preparation of Caffeine-Benzo-15-Crown-5 Conjugate
The caffeine conjugate was prepared according
to step 4 in the reaction scheme except that the
.
125q331
-56-
protected l-methyl xanthine was replaced with
1,7-dimethyl xanthine.
To 50 ml of DMF in a 125 ml conical flask was
added l.Og (5.56 mmole) of 1,7-Dimethylxanthine.
The mixture was warmed with stirring until dis-
solution occurred. The solution was cooled to room
temperature and 0.59g of anydrous sodium carbonate
was added followed by the addition of 2.95g (5.56
mmole) of the previously described crown ether amide
and the reaction mixture stirred for 3 days. The
DMF was removed by rotary evaporation and 100 ml of
water was added and the mixture extracted 3 times
with 50 ml of chloroform. The combined extracts
were washed 1 time with 100 ml of saturated aqueous
sodium chloride and dried over anhydrous sodium
sulfate. Evaporation of the chloroform yielded a
brown oil which was crystallized from hexane-ethyl
acetate to yield l.Og (29%) of product. The product
was recrystallized from ethyl-acetate/hexane with
charcoal clarification and found to be homogenous by
hplc. m.p. 130-133. Analysis calculated for
C32H47N5O8: C:61.03; H:7.52; N:11.12. Found:
C:60.78; H:7.53; N:10.97. The benzo-12-crown-4
derivative was prepared by a similar method.
PreParation of A Drug Specific Electrode for The
Measurement of Theophylline
Theophylline Membrane
The components for the theophylline membrane
consisted of: 0.005 g theophylline-crown ether
conjugate, 0.02 g potassium tetra(p-chlorophenyl)
borate, 0.5 ml bis(2-ethylhexyl) sebacate, 0.3 g
~2S7331
-57-
high molecular weight PVC and 5 ml THF. The mem-
brane can be formed by either solvent casting or dip
coating as previously described.
Caffeine Membrane
The components for the caffeine membrane
consisted of: 0.005 g caffeine-crown ether conju-
gate, 0.02 g potassium tetra(p-chlorophenyl) borate,
0.5 ml bis-(2-ethylhexyl) sebacate, 0.3 g high
molecular weight PVC and 5 ml THF. The membrane can
be formed by either solvent casting or dip coating
as previously described.
Differential Measurement Technique
One membrane each of theophylline and caffeine
was mounted in an Orion 95 electrode body. The
internal filling solution consisted of pH 7.4
phosphate buffered saline and the reference elec-
trodes were silver-silver chloride wires. The two
electrodes were fitted into a rubber stopper and
immersed into 5.0 ml of pH 7.4 phosphate buffered
saline. The solution was magnetically stirred. The
emf measured between the two electrodes was 0.23 mv.
To test the response to potassium ion, 100 ul of 0.2
molar KCL was pipetted into the stirred buffer solu-
tion. A transient response was observed as evi-
denced by a sharp peak (0.3 mv) on the output of astrip chart recorder. The pGtential difference
returned to the base-line value within 30 to 60 s.
The transient response to potassium is probably due
to a local concentration of potassium ion near the
membrane surface which disappears as the solution is
mixed.
~2573:~1
-58-
To test the response to protein, 100 ul of a
- standard protein solution (Sigma Chemical Co., St.
Louis, MO, Cat. # 540-10) consisting of 3.5 g/dl
albumin and 3.0 g/dl globulin was added to the
solution. The standard protein solution had no
effect on the electrode potential.
When, however, 100 ul of theophylline antibody
(Immunotech Corp., Allston, MA, Cat. # 651 or
Research Plus, sayonne, NJ, Cat. # 01-9603-09) was
added the potential difference between the two
electrodes shifted by 0.85 mv over a period of about
20-30 min.
A sensor of this type can be used in a competi-
tive binding assay. The antibody which remains
unbound in the sample is measured and provides an
indirect measure of antigen present in the sample.
The more antibody bound by the antigen in the
sample, the lower the resulting electrical signal
because less antibody will be available for binding
to the antigen immobilized in the sensor membrane;
that is, antigen concentration will be inversely
related to the magnitude of the signal.
Example 8 PreParation of an Ion Selective Electrode
for the Measurement of Creatinine
Creatinine is an important indicator of renal
function and its measurement in serum is a routinely
performed blood test. Presently, in the clinical
laboratory there are two ways of determining cre-
atinine. The more widely used method is the Jaffe
reaction, which is based on the production of a red
colored complex between creatinine and picrate in
~ZS73:~
-59-
alkaline solution. This method gives erroneous
results in the presence of certain metabolites and
drugs. The second method utilizes an enzyme.
Although enzymic methods are highly specific, the
expense and the time required to carry out a test
restrict their routine use in clinical laboratories.
Most recently a creatininium ion selective
electrode has been descrbed. E.P. Diamondis and
T.P. Hadjiiannou, Anal. Lett., 13, 1317-1332 (1980).
The creatinium ion exchanger was prepared by mixing
equimolar aqueous solutions of creatinine and sodium
tetraphenylboron followed by the addition of a
hydrochloric acid solution which results in precipi-
tation of the complex salt creatininium tetra-
phenylboron. The salt was extracted into 2-nitro-
toluene and this solution of the complex salt used
as a liquid ion exchanger in an Orion 92 (Orion Re-
search, Cambridge, MA) electrode equipped with
Teflon membranes. In this type of ion selective
electrode the ion exchange material is in liquid
form contained in a reservoir and continuously
leaches through the membrane into the sample solu-
tion and therefore must be periodically replenished.
The electrode was conditioned by soaking in 0.01
mol/L creatininium chloride for 24 hours before use.
The response of the creatininium electrode at pH 3
was linear in the 10 to 10 mol/L range with a
slope of 57 mv at 20C. The slope decreased to 37
mv in the 10 4 to 10 3 mol/l range.
The present invention makes use of a plastic
membrane in which the ion exchange material is
~2S733~
-60-
immobilized. When creatininium tetraphenyl boron
was incorporated into a PVC membrane using 2-nitro-
toluene as a plasticizer a slope of 44 mv was ob-
tained between 10 3 and 10 1 mol/L creatininium
chloride and decreased to 14 mv in the 10 4 to 10 3
mol/L range. The clinically important range for
creatinine is 10 to 10 mol/L. The use of other
nitrated plasticizers such as 2-nitro-p-cymene and
2-nitrophenyloctyl ether did not improve the sensi-
tivity in the 10 4 to 10 3 mol/L range.
A significantly improved response in the 10
to 10 3 mol/L range was obtained when a substituted
tetraphenylboron salt was used having the general
formula:
( ~B (H ~
! C~3
where x = Cl, Br, F, CF3
In one embodiment, a membrane was prepared consist-
ing of 0.02 g of creatininium tetra-(p-chlorphenyl)
boron, 1.5 ml of 2-nitrophenyloctyl ether, 0.3 g of
very high molecular weight PVC, and 5 ml of THF.
The membrane was prepared either by solvent casting
or by dip coating as previously described. The
membrane yielded a slope of 55 mv between 10 3 ml/L
and 10 1 mol/L creatininium chloride and 40 mv in
the 10 4 to 10 3 mol/L range and did not require any
preconditioning. The substituted tetraphenyl boron
;~ salts exhibited much higher sensitivity in the
clinical range of concentration than the unsubsti-
tuted tetraphenyl boron salt. The performance of
lZ57~3~
-61-
the various substituted tetraphenyl boron salts is
summariæed in Table 2.
The unexpected increase in sensitivity of the
substituted tetraphenyl boron salts over the un-
substituted tetraphenyl boron may be attributed toan increase in lipophilicity and differences in the
nature of the ionic exchange properties of the
complexes formed with the substituted tetraphenyl
boron salts. In practice, the pH of the solutions
containing creatinine must be decreased to a pH of 3
or less by the addition of acid. This can be
accomplished by dilution of the sample with an acid
solution. However, in the case of a serum sample,
significant dilution may adversely effect the
sensitivity of the assay due to a decrease in the
concentration of creatinine.
Tn the present case a different approach was
taken. The pH of the sample containing creatinine
was lowered by treating the sample with a support
2Q containing dried glycine hydrochloride. This was
accomplished by dipping a membrane support such as
nitrocellulose (Schleicher and Schull 12 micron,
grade HE-100) into an aqueous solution of glycine
hydrochloride (0.1 to 1.0 moles/L). A sandwich was
formed such that the nitrocellulose membrane covered
the surface of the creatininium tetra (p-chloro-
phenyl) boron PVC membrane. Referring to Figure 4k,
each cell a, b and c contains a creatininium PVC
membrane and a nitrocellulose membrane containing
glycine hydrochloride. The serum sample is added to
c and reference and calibrant solutions are added to
~'25~
-62-
a and b, respectively. The glycine hydrochloride
converts the creatinine to creatininium hydro-
chloride which is detected by the membrane.
The emf developed between a and b may be used
to calibrate the sensor. The slope from a and b and
the emf developed between b and c may then be used
to determine the creatinine concentration in the
sample.
Preparation of Creatininium Tetraphenyl soron Salts
The creatininium tetraphenyl boron salts were
prepared by mixing equimolar aqueous solutions of
creatininium hydrochloride (Sigma Chemical Co., St.
Louis, MO, Cat. # C-6257) and the corresponding
sodium salt of the substituted tetraphenyl boron.
The precipitated salts were collected by filtration,
washed with water and dried. The sodium salts of
the substituted tetraphenyl boron derivatives were
prepared by the method of Cassoretto, McLafferty and
Moore, Anal. Chem. Acta, 32:376-380 ~1965) and
involved the following steps:
X ~ B~
>< ~ ~ g~ F~
~_ ~,Cl,CF3
125~7331
-63-
Table 2. Response of Creatininium Tetraphenyl Boron
Salts to Creatininium H~drochloride.
X 10to 10 mol/L 10to 10 mol/L
-
Slope(mv) Slope(mv)
H 50 10
F 45 26
Cl 55 36
CF3 57 40
All mv readings shown in Table 2 are versus a double
junction silver-silver chloride reference electrode.
The creatininium solutions used were prepared with
0.05 M glycine-glycine hydrochloride buffer pH 3.0
containing 150 mM sodium chloride and 6 mM potassium
~ 15 chloride.
; Example 9 PreParation of an Ion Selective Electrode
for the Measurement of Sodium
-
A membrane selective to sodium ions was pre-
pared using the following proportions of components:
20 0.2g methylated monensin, 0.04g sodium tetra-
(p-chlorophenyl) borate, 2.0 ml 2-nitrophenyloctyl
ether, 1.2g high molecular weight PVC, 12 ml
tetrahydrofuran (THF) and 5 ml 4-methyl-2-pentanone.
The membrane can be formed by either solvent casting
or dip coating as previously described.
The membrane was incorporated into the
previously described electrode housing and tested
using aqueous sodium ion containing solutions. The
sensor yielded a slope of 57-60 mv per decade at
25.
1'2S73~3~
-64-
Preparation of Methylated Monensin
- To 250 ml oi 1,4-dioxane in a 500 ml round
bottom flask was added 5.0 g of sodium monensin
(Sigma Chemical Co., St. Louis, MO, Cat. # M2513)
along with 20 ml of iodomethane. The mixture was
gently refluxed for 18 hours and then evaporated to
dryness on a rotary evaporator. The residue was
dissolved in 200 ml of methylene chloride and washed
in turn 2 times with 500 ml saturated aqueous sodium
chloride, 2 times with 200 ml 5% aqueous sodium
hydrogen carbonate and one time with 200 ml of
saturated aqueous sodium chloride. The methylene
chloride solution was dried over anhydrous sodium
sulfate and evaporated to dryness. The residue was
treated with 50 ml of hexane with warming in a 50
water bath. The hexane solution was cooled in a
freezer for 1-2 hours followed by filtration to
remove precipitated salts. The hexane solution was
evaporated to dryness to yield 2.5-4.0 g of a light
tan oil. The oil was used to prepare the sodium
selective membrane.
Example 10 Preparation of an Ion Selective
Electrode for the Measurement of
Potassium
A membrane selective to potassium ions was
prepared using the following proportions of com-
ponents: 0.2 g valinomycin, 0.04 g potassium tetra-
phenyl borate, 2.0 ml 2-nitrophenyloctyl ether, 1.2
g high molecular weight PVC, 10 ml THF and 5 ml
4-methyl-2-pentanone. The membrane can be formed by
either solvent casting or dip coating as previously
described.
1257331
--65--
Example 11 Preparation of an Ion Selective
Electrode for the Measurement of
Chloride
A membrane selective to chloride ions was
prepared using the following proportions of com-
ponents: 1.0 ml Aliquat 336 (Aldrich Chemical Co.,
Milwaukee, WI, Cat. # 20,561-3), 2.0 ml tris
(2-butoxyethyl) phosphate (Aldrich Chemical Co.,
Milwaukee, WI, Cat. # 13,059-1), 1.2 g very high
molecular weight PVC (Aldrich Chemical Co., Mil-
waukee, WI, Cat. # 18,261-3), 12 ml THF and 5 ml
4-methyl-2-pentanone. The membrane can be formed by
either solvent casting or dip coating as previously
described.
Industrial Utility
This invention has industrial utility in the
determination of the ion content or concentration of
other constituents of samples, in particular
biological fluids such as blood, serum, plasma,
urine, saliva and cerebrospinal fluid. It is
particularly useful for the rapid determination of
the ion activity of a biological sample, as well as
the concentration of other sample components, such
as glucose, urea, triglycerides, creatinine, uric
acid, lipase, other enzymes and drugs. It is well
suited for use in a clinical or research context
because there is no need for preconditioning of the
ion-selective membrane and because it provides
results quickly. In addition, the invention can be
- r~
~5~331
-66-
used for the similar determination in other samples
such as beverages, meats, canned and processed
foods, fruit extracts, etc.
Equivalents
Those skilled in the art will recognize, or be
able to ascertain using no more than routine experi-
mentation, many equivalents to the specific compo-
nents and materials described specifically herein.
Such equivalents are intended to be encompassed in
the scope of the following claims.
