G. M. MURRAY ET AL.

APPLIED RESEARCH

A

Molecularly Imprinted Polymers for the SelectiveSequestering and Sensing of Ions

George M. Murray, Amanda L. Jenkins, Anton Bzhelyansky, and O. Manuel Uy

variety of polymeric materials were developed as the basis for sequestering andsensing ions in aqueous solution. Polymers were made that can act as selective ionexchange resins. The selectivity can sometimes be made to approach exclusivity andallow the simultaneous removal and purification of a variety of chemicals. Thisapproach can be used to turn waste streams of hazardous chemicals into resources ofvaluable materials. In addition, some of the polymers were used to construct ionselective electrodes for sensing ions in aqueous solution. Similar polymers wereconstructed with complexing monomers that possess chromophores, allowing for theconstruction of chemically selective optical ion sensors. The result of these studies isa general approach for the construction of selective ion sequestering and sensingpolymers.(Keywords: Ion exchange, Molecularly imprinted polymer, Sensors.)

INTRODUCTIONAn ideal reagent for use in constructing ion selective

sequestering agents and sensors would be one for whichonly a specific ion would be complexed. An ideal ionselective separation or measurement technique wouldtake place in a simple single-stage operation involvingno elaborate apparatus, be rapid, require little or nopretreatment of the aqueous phase (required for in situmeasurements), involve an agent that is insoluble inthe aqueous phase, and entail minimal cost. This re-search is aimed to develop polymeric complexingagents that meet these criteria.

In nature, selective complexing agents are made bythe immune system. Molecularly imprinted polymers(MIPs) are synthetic materials designed by emulatingthe way the immune system generates the selectivecomplexants known as antibodies. Specifically, MIPstrace their origin to suppositions about the operationof the human immune system by Stuart Mudd in the

464 JOH

1930s and Linus Pauling in the 1940s. Mudd’s contri-bution to understanding the immune system was topropose the idea of complementary structures.1 Thissupposition states that the reason that a specific anti-body attacks a specific target or “antigen” is because theshape of the antibody provides an excellent fittingcavity for the shape of the antigen. The description isvery similar to the “lock and key” analogy used toexplain the action of enzymes, the molecules respon-sible for hastening and directing biochemical reactions.Enzymes form a lock for a particular chemical key tofit, and as the “key” is turned, the enzyme directs andhastens the production of desired products from thechemical target.

Pauling’s contribution to the development of MIPswas to explain the source of the complementary shapeexhibited by antibodies.2 He postulated how anotherwise nonspecific antibody molecule could be

NS HOPKINS APL TECHNICAL DIGEST, VOLUME 18, NUMBER 4 (1997)

MOLECULARLY IMPRINTED POLYMERS

reorganized into a specific binding molecule. He rea-soned that shape specificity was obtained by using atarget antigen to arrange the complementary shape ofan antibody. Thus, a nonspecific molecule can beshaped to the contours of a specific target, and whenthe target is removed, the shape is maintained to givethe antibody a propensity to rebind the antigen. Thisprocess is known as molecular imprinting or “templating.”

The target or template molecule directs the position-ing of the encapsulating antibody by the interactionsthat occur between certain sites on the target andcomplementary sites on the antibody. The sites thatallow complementary associations are certain arrange-ments of atoms that exhibit an electrostatic attractionof a specific kind. These localized atomic arrangementsare called “functional groups” by chemists. The function-al groups on a molecule help to define the molecule’soverall chemical properties. Functional groups mayform attachments to their complements by covalent,ionic, or hydrogen bonding. In the context of ion-exchange materials, the association is through ionicbonds.

The early work in molecular imprinting was basedon silica gels,3,4 which showed some retention or mem-ory of structure but were limited in application. Thesynthesis and characterization of molecularly imprintedorganic polymers were first reported by Wulff andSarhan5 in 1972. Their investigations, and later thoseof Shea and others,6,7 and Damen and Neckers,8–10

indicated that polymers with specific cavities can beproduced.

To be useful, Wulff and Sarhan5 indicated that thesepolymers must exhibit the following characteristics:

• The cavity geometry and the binding group arrange-ment should be preserved after the removal of thetemplate molecule. This may be accomplished byhigh degrees of cross-linking, that is, use of a structuralmonomer that links chains together.

• The cavities must retain enough flexibility to permitrapid uptake and release of the template molecule.This property decreases with increasing cross-linking,so a balance must be sought.

• A large fraction of the template molecules should beremovable after polymerization. This implies goodaccessibility of the cavities, which requires polymerswith large surface areas.

• The polymeric materials should be chemically andmechanically stable to permit regeneration and reuse.

Wulff and Sarhan5 and the others6–10 also suggested twobasic approaches to the preparation of site-specificpolymers: (1) polymerize monomers that contain bind-ing groups with set geometries by virtue of attachmentto a template molecule, and (2) produce binding groupslocated on a flexible polymer chain, and then fix them

JOHNS HOPKINS APL TECHNICAL DIGEST, VOLUME 18, NUMBER 4 (1

in a definite geometry by attachment to a templatemolecule followed by cross-linking.

The production of polymers with selective bindingfor a specific cation is achieved by providing polymerswith cavities lined with complexing groups or “ligands”arranged to match the charge, coordination number,coordination geometry, and size of the target cation.Anion complexing polymers are made in a similar man-ner, but typically employ a trapped metal ion that hasa large affinity for the anion in question. These cavity-containing polymers are produced by using a specificion as a template around which monomeric complexingligands will be self-assembled and later polymerized.

The complexing ligands are ones containing func-tional groups known to form stable complexes with thespecific ion and less stable complexes with other ions.The process is outlined in Fig. 1. We have extended andamended previous methods in our work, as detailedlater, by changing the order of the steps, by the inclu-sion of sonication, by using higher template complexloading, and by the selection of functional groups withbetter complexation constants. Further, we have usedthe imprinted polymers as selective extractants and infabricating selective ion sensors.11–13

POLYMER PRODUCTIONThe key step in making a molecularly imprinted

polymer is to form a complex that will survive thepolymerization process and leave behind a suitable setof binding sites when the templating species is re-moved. To form such a complex, ligands must be chosenthat exhibit sufficiently large affinities to resist disso-ciation. The polymerization process must provide suf-ficient rigidity to effect structural “memory” but besufficiently flexible to allow removal of the templateion. In the case of sensors, the ligating monomers mustprovide a means of signal transduction. The success ofthe end product hinges on the selection of the ligatingmonomer.

Ligand Monomer SelectionInitially, monomeric ethylene, propylene, and sty-

rene derivatives with substituted coordinating groups(X), both acidic and neutral (H2C=CH-X, H2C=CH-CH2-X, and H2C=CH-C6H4-X), were evaluated asspecific ion complexing agents. Some were purchasedand others were synthesized. Preparations for thosecompounds not available commercially are well knownand readily available in the literature. Ligands wereselected on the basis of the thermodynamic affinities forthe specific metal ion (maximum affinity) versus theaffinities for competing metal ions (maximum differ-ence in affinities). Affinity data are readily available

997) 465

G. M. MURRAY ET AL.

Figure 1. Schematic representation of molecular imprinting. 1,2: Self-assembly of the templates and ligand molecules; 3: incorporationof the ligand–template complex into the polymer matrix; 4: removal of the template molecule; 5: formation of the templated cavity.

1 2

3

4

5Template molecule

Ligand molecules with vinyl groups

from the chemical literature, for example, the tables ofcomplexation constants in the CRC Handbook, Gmelin,and other sources.14,15

On the basis of the affinities information, vinylben-zoic acid was chosen as the first candidate ligand forlead(II) ion complexation. Although vinylbenzoic acidis commercially available, we found that it requiredextensive purification before use. Therefore, it wassynthesized in-house by the Wittig reaction. Severalother vinyl-substituted ligands were also used. Methyl-3,5-divinylbenzoate was synthesized by the method ofShea and Stoddard.16 The divinyl compound was pre-pared primarily to give greater structural integrity to thecoordination site by providing each ligand with twopossible connections to the polymer network. It wasanticipated that the methyl-esterified compound wouldeliminate the pH dependence, characteristic of thevinylbenzoic acid carboxyl functionality. 5-Vinylsalicyl-aldoxime was synthesized by a method adapted fromWulff and Akela17 and tested as a selective agent forboth lead(II) and uranyl ions.

Synthesis of Complexes with PolymerizableLigands

With a careful choice of complexing ligands, metalcomplexes were synthesized by mixing stoichiometricamounts of a metal salt and the complexing ligand inaqueous solution and evaporating to near dryness. Ourexperience with making complexes has shown thatwater or alcohol/water mixtures of the metal and ligandin stoichiometric ratios, evaporated to dryness, result in

466 JOH

near quantitative yield of the desired complex com-pound. The stoichiometry of the complexes was veri-fied by determining the mass percent of metal ion usinginductively coupled plasma atomic emission spectros-copy (ICP-AES). To make complexes that containtarget anions, it was necessary to make mixed ligandcomplexes that have a one-to-one stoichiometric ratioof target anion to complex. This was done by synthe-sizing metal ion complexes with the proper coordina-tion number of tightly binding ligands such that a singletarget analyte could bind by replacing a very weaklybound substituent. This step can be difficult if insuffi-cient care is taken in selecting the appropriate ligands.

Synthesis of the CopolymersHere and elsewhere in this article, cross-linking is

defined as the mole percent content of the cross-linkingagent, divinylbenzene. The polymer matrix was initial-ly formed at a relatively high cross-link level to enforcethe retention of the cavity “memory” upon the removalof the metal ion. Experience with templated resins formetal ions that exhibit directed bonding has sincerevealed that relatively small amounts of cross-linkingresult in more selective resins. This may be a result ofthe bulk of the polymer possessing metal ion-to-ligandionic and dative bonds as well as covalent cross-links.

The introduction of microporosity for increasedreagent accessibility was achieved through the additionof a noncoordinating monomer (styrene). Solutions ofthe monomeric complex (metal ion complexed withpolymerizable ligands), styrene, initiator, and cross-

NS HOPKINS APL TECHNICAL DIGEST, VOLUME 18, NUMBER 4 (1997)

MOLECULARLY IMPRINTED POLYMERS

linker (divinylbenzene) in various amounts of a suitablesolvent (0–20 weight percent of pyridine) were pre-pared. The dissolution of the complex/monomer mix-ture has historically been a difficult step. We havefound that this step can be simplified if the copolymer-ization is carried out using ultrasonication. The soni-cation has the serendipitous effect of maintaining thetemperature at about 60°C, which is the recommendedtemperature for chemically initiated free-radical poly-merization using azobisisobutyronitrile. Lead vinylben-zoate copolymers were produced with and withoutsonication, and it was determined that the sonicationresults in a more homogeneous product but with onlya slight increase in capacity over the polymers obtainedby heating in an oil bath. On the other hand, exam-ination of some uranium-containing copolymers hasshown larger improvements (greater than 10-fold) incapacity due to sonication. The expectation was thatin addition to improving solubility, the sonicationwould result in more complete incorporation of themetal ion complex in the copolymer. It is apparent thatthe effectiveness of sonication varies with the choiceof metal ion complex and may or may not be simplya function of miscibility.

When the polymerization was completed, the cross-linked polymer was washed, cryogenically ground toa uniformly fine powder, and extensively eluted withnonpolar solvents to remove unreacted complex.A grinding step was necessary to maximize surface areaand allow for access by the various reagents andsamples. It required freezing the polymer in liquid ni-trogen and using a ball mill employing a stainless steelcapsule. This freezing made the polymer brittle enoughto be ground and prevented distortions of the polymerby the heat of friction. Sieves were used to obtain theuniformly sized particles. Polymers used in the con-struction of optical sensors were prepared in situ on thedistal end of an optical fiber whose surface was preparedby binding a polymerizable agent on the surface.

Acetone or another organic solvent was used toswell the polymers, allowing greater access to the co-ordinated metal ions. This step was necessary since,unlike commercial resins whose large amount of func-tionalization (ionogenic sites) causes the resin to swellmost in polar solvents, the templated resins have arelatively low amount of functionalization and areprimarily nonionic matrices. The template moleculecan be removed in a variety of ways. For example, formetal ion templated resins, subsequent to the removalof unreacted monomer, a 1 N aqueous acidic solutionwas mixed into the acetone washes, with increasingaqueous acidic phase in each sequential wash, to re-move the metal ions from the cavities by mass action.A significant fraction of the cations (but not all) wasremoved, thus leaving cavities that were cation-specific.Those cations that were not removed even after

JOHNS HOPKINS APL TECHNICAL DIGEST, VOLUME 18, NUMBER 4 (1

swelling were locked inside the polymer in inaccessiblesites. Cations can be removed from fiber-bound poly-mers with the aid of a chelating agent such as ethyl-enediaminetetraacetic acid (EDTA) after swelling in anappropriate solvent. Acidity is avoided in this case sothat the polymer-to-silica bond is not hydrolyzed.Anions may also be removed by mass action withacidification when the anion can be protonated, leav-ing in its place the conjugate base of a strong acid suchas chloride or nitrate.

ANALYTICAL APPLICATIONSOF THE METAL ION TEMPLATEDPOLYMERS

The ion templated polymers are being used in avariety of analytical applications. Lead selective resinswere applied to the analysis of seawater in a head-to-headexperiment with three other ion selective resins. A fastsensitive method for determination of lead in seawaterwas developed. Lead(II) and uranyl ion selective mem-brane electrodes were fabricated using ground polymerparticles as sensing elements. Numerous selectivity stud-ies were conducted to characterize the selectivity of thesensors. A fluorescent lead templated polymer is beingdeveloped as an optical sensor (optrode) for determina-tion of the lead(II) ion concentration in aqueous solu-tions. And finally, a polymer templated for the hydrolysisproduct of the nerve gases sarin and soman has beendeveloped as the basis of a portable sensor for thesesignificant chemical threats.18

Templated Polymers as Sequestering AgentsThe first application was a comparative study of ion

exchange materials for the extraction and preconcen-tration of the lead(II) ion prior to analysis by ICP-AESand inductively coupled plasma mass spectrometry(ICP-MS). The lead(II) ion was extracted from seawa-ter by employing four different ion exchange resins:Chelex-100, a proprietary NASA resin, our lead tem-plated vinylbenzoic acid resin, and a commercial thiol-based ion exchange resin. Seawater is one of the mostchallenging matrices for chemical analysis and waschosen because of the difficulty it poses in chemicalanalysis. Measurement of the lead(II) concentrations,as well as the concentrations of possible interferingspecies, was performed first with ICP-AES and laterwith ICP-MS. Recoveries of lead(II) were determinedto evaluate the suitability of the various resins as se-questering agents and to show the advantage in selec-tivity of the templated ion exchange material.

The results of this experiment exceeded expectations.The preconcentrated lead extracts obtained using thetemplated resin were found to be exceedingly pure. Be-cause of the relative purity (absence of significant

997) 467

G. M. MURRAY ET AL.

amounts of interferents) of the seawater extract ob-tained from the templated resin, it was tested for leadby a simple photometric method that normally requiressignificant purification and preparation. This quantita-tive analysis was performed using dithizone complex-ation with spectrophotometric detection (measuringabsorbance at 522 nm). Dithizone is a colorimetricreagent that forms colored complexes with a wide rangeof metal ions. The results of the photometric analysisagreed with the ICP-AES data. The use of the templat-ed resin in this manner provides an inexpensive assayfor the lead(II) ion in aqueous solutions with a detec-tion limit of <100 ppb and a linear range from 100 ppbto 2.00 ppm. The use of the resin with larger volumesof analyte would allow for preconcentration and muchlower limits of detection. The detection limit of 100ppb is already better than the normal ICP-AES detec-tion limit of about 1 ppm.

The limitation on the use of the templated resinsfabricated by the preceding methods is the relativelysmall exchange capacity. The low capacity is due to thebulk of the ionogenic sites being buried in the polymerparticle and thus being unavailable for exchange. Inpart, this handicap is overcome by the selectivity of theresin. Since the resin is so selective for the metal ofchoice, few of the available sites are occupied by inter-ferents, and so less capacity is required for most appli-cations. Another consequence of the activity of theresin being restricted to the surface of the particle is therapidity of ion extraction, which allows for higher flowrates for extractions and is particularly attractive foranalytical preconcentration.

Templated Polymers in Ion Selective ElectrodesAfter analyzing the sorbent properties of the metal

ion templated polymers, the polymers were used tofabricate ion selective electrodes. Ground polymer ofan appropriate size, 60 mesh, was cast in a polyvinylchloride membrane. The membrane was adhesivelybonded to a polyvinyl chloride support tube and assem-bled into a standard polymer membrane electrode. Theselectivity and stability of the templated polymer givethis electrode advantages over traditional ion selectiveelectrodes. An extensively used electrode does notdisplay a loss of function resulting from solubility of theactive ingredient, and the electrodes are immune to thechemical poisoning that occurs with crystal-based elec-trodes. A typical response curve for the templated poly-mer electrode for the lead(II) ion is shown in Fig. 2.

The potential responses of the electrode to a varietyof common interference ions are shown in Fig. 3. Asexpected, the electrode based on the lead(II) ion im-printed polymer membrane showed a strong preferencefor lead(II) ions. Membranes based on a 3 mole percentequivalent vinylbenzoic acid polymer and on benzoic

468 JO

acid were also prepared and examined. The membranebased on the lead templated polymer was found to bemore sensitive than either the membrane based on theuntemplated polymer or the one prepared with dis-solved benzoic acid.

The production of polymers for use with other metalions has begun with templating a polymer for the uranylion. Figure 4 shows the response of a membrane pre-pared with a uranyl ion templated polymer. In this case,the result was astounding in that no interference hasyet been detected. It was thought that the unique shapeof the uranyl ion would make it a prime candidate forthe template process. This expectation has been real-ized. The influences of unique shape and other factorsare currently being investigated for uranyl templatedpolymers.

Templated Polymers in Optical Sensing

Lead Ion Optical Sensor

During the initial screening of coordinating mono-mers for possible ion selective electrode enhancements,

180

160

140

120

100

80

–6 –5 –4 –3 –260

Pot

entia

l (m

V)

Log lead(II) activity

Figure 2. Calibration curve for a lead(II) ion selective electrodeconstructed with a templated polymer cast in a polyvinyl chloridemembrane (slope = 31.7, correlation coefficient = 0.996).

200

180

160

140

120

100

80

–6 –5 –4 –3 –2

Log metal ion activity

Pb Ca Cd Zn Cu Na La

Pot

entia

l (m

V)

Figure 3. Response of a lead(II) ion templated electrode and themost likely interferent ions.

HNS HOPKINS APL TECHNICAL DIGEST, VOLUME 18, NUMBER 4 (1997)

MOLECULARLY IMPRINTED POLYMERS

it was observed that methyl-3,5-divinylbenzoate(DVMB) luminesces blue, and that the luminescencebecomes yellow–green when a lead(II) complex isformed. This observation suggests phosphorescence dueto an external heavy atom effect. Since the lead(II)–DVMB complexes rapidly polymerize in solution, it wasdecided to investigate the optical properties of theligand precursor, methyl-3,5-dimethylbenzoate. Thelead(II) complex was prepared by the method outlinedpreviously and dissolved in hexane to prevent complexdissociation. Luminescence spectra of the lead(II) com-plex and free ligand were obtained with a spectroflu-orimeter using a 467-nm excitation wavelength. Serialdilutions were used to obtain analytical figures of meritfor a solution assay of the lead(II) complex in hexane.The calibration curve was linear for four decades ofconcentration from 70 to 7.0 × 105 ppb complex witha limit of detection of 50 ppb complex (equivalent to20 ppb lead), having a correlation coefficient of 0.997.

These observaobtained with a vaed convincing evcould be made. Thdivinylbenzene) won a vinylized 400was then used in Fig. 5. The lumicoated optical fibthe lead(II) comdisplayed in Fig. 6from the polymermethanol and wasolution of EDTAluminescence specer exhibited the cnescence band.

The standard coptrode is given icited by the 488-nof about 10 mW. Tity and interferenctrials in an envirosources of wastew

Counter-Terrorism

The sensor syssigned to measuresarin and soman.agents degrade upproducts of hydrodesign allows presensor surface by Also, the use of a obviates the nee

250

200

150

100

–6 –5 –4 –3 –2Log metal ion activity

UO22+

Pb Cd Ca Cu Zn Ni Na La

Pot

entia

l (m

V)

Figure 4. Response of a uranyl ion templated electrode (slope =29.0, correlation coefficient = 0.995) and the most likely interferentions.

Figure 5. Prototype of a lead(II) ion measuring optrode system. (A) argon ion laser, (B)dichroic mirror, (C) mirror, (D) lens, (E) fiber mount, (F) optical fiber core, (G) fiber cladding,(H) templated polymer coating, (I) sample, (J) monochromator, (K) detector, (L) analog-to-digital converter, (M) computer.

KJ

B C

A

M

I

E

D

L

F

G

H

JOHNS HOPKINS APL TECHNICAL DIGEST, VOLUME 18, NUMBER 4 (

tions, along with similar resultsriety of complexing ligands, present-idence that a lead sensing optrodee 3% lead(II)–DVMB complex (2%as bound by in situ copolymerization-mm optical fiber surface. The fiber

the setup schematically presented innescence spectrum of the polymer-er showed the characteristic band ofplex, virtually identical to the one. The lead(II) ion was then removed by first swelling it in a mixture ofter and then soaking it in a stirred; each process took about 1 h. Thetrum of the cleaned polymer no long-haracteristic lead(II) complex lumi-

urve generated by the first prototypen Fig. 7. The luminescence was ex-m line of an argon ion laser at a powerhe device is being tested for longev-e effects and is to be applied to field

nmental project aimed at identifyingater seepage into streams.

Nerve Gas Sensor

tem being developed has been de- the primary hydrolysis products of This is the process by which theon exposure to moist air, and thelysis are relatively innocuous. This

concentration of the analyte at theenhancing analyte-to-sensor affinity.hydrolysis product of the nerve agentd for exposing personnel to toxic

chemicals during the developmentof the device.

The laboratory prototype for thesarin/soman fiber-optic sensor (Fig.8) consists of an optical fiber ontowhich the polymer has been coat-ed. The fibers were prepared by ter-minating one end of a 400-mmoptical fiber with a standard opticalfiber connector and removing thecladding from the distal end. Thevinyl-silanized end of the fiber wasthen dip-coated with a viscouspolymer solution, leaving a uni-form coating on the fiber. Thepolymer was then cured under ahandheld UV lamp overnight.Once cured, the polymer on thefiber was cleaned using the swellingand removal procedure describedpreviously.

1997) 469

G. M. MURRAY ET AL.

1.2

1.0

0.8

0.6

0.4

0.2560540520500480

Wavelength (nm)

Methyl-3,5-dimethylbenzoate

Pb(methyl-3,5-dimethylbenzoate)2Br2R

elat

ive

inte

nsity

Figure 6. Luminescence spectra of a free ligand (methyl-3,5-dimethylbenzoate) and its lead(II) complex excited at 467 nm.

2000

1500

1000

500

0800600400200

Concentration (ppb)

Rel

ativ

e in

tens

ity

Figure 7. Calibration curve obtained for the first prototype lead(II)optrode (slope = 2.34, correlation coefficient = 0.9994).

The prototype sensor coated with the 3 mole percentcomplex polymer is currently being tested. Lumines-cence is being excited using the 488-nm laser line withthe response carried to the monochromator through the

Figure 8 . Laboratory prototype for the sarin/soman fiber-optic sensor.

fiber. Thus far, the sensor has been able to detect thehydrolysis product at levels from 150 ppm to 1 ppb witha linear response (Fig. 9). The limit of detection for thisinitial sensor is 125 parts per trillion of pinocolyl

methyl phosphate in 0.1 M sodiumhydroxide. The response time ofthe sensor is typically less than 1min. Studies on limit of detectionand interferences are currently be-ing performed.

Once the preliminary testing iscompleted, the working end of thesensor will be enclosed in a semi-permeable membrane with an alka-line solution using a solvent such aspolyethylene glycol to preventevaporation. (The alkaline solu-tion is necessary to catalyze thehydrolysis of the nerve agents totheir respective hydrolysis prod-ucts.) Once the prototype is com-pleted, it will be subjected to test-ing with the actual nerve agents atan approved site. The portabledevice will be designed using asmall photosensor module for de-tection, with the output going to agated integrator. An optical multi-plex switch will be incorporatedinto the design so that many sen-sors can be coupled to one controlsystem. This setup should allowmonitoring of a large building, sub-way station, or airport.

CONCLUSIONSMolecular templating is an

emerging technology for the cre-ation of selective binding sites in

470 JOHNS HOPKINS APL TECHNICAL DIGEST, VOLUME 18, NUMBER 4 (1997)

MOLECULARLY IMPRINTED POLYMERS

Figure 9. A log–log calibration plot for the sarin/soman sensor(correlation coefficient = 0.999).

6

5

4

3

2

1

0543210

Log concentration (ppb)

Log

area

synthetic polymers. Ion selective sensors have theadvantages of being simple, rapid, and inexpensive.This research explores the utilization of metal ion tem-plated polymers as selective sequestering agents and ascomponents of electrochemical and optical sensors.Results from the characterization of the metal ion tem-plated electrodes for lead(II) and uranyl ions demon-strate a broad linear range of measurement with a near-Nernstian response (the theoretically expectedresponse according to the Nernst equation) and astrong preference for the templated metal ion over theother metal ions tested. The lifetime of the electrodeis relatively long, which is one of the advantages ren-dered by templated polymers with the functional groupscovalently attached to the polymer matrix. Opticalsensors based on templated polymers have been shownto determine the lead(II) ion and a nerve gas hydrolysisproduct in the parts per billion range. The informationobtained from these studies will be used to produce avariety of new ion selective polymers and to fabricateadditional sensors.

JOHNS HOPKINS APL TECHNICAL DIGEST, VOLUME 18, NUMBER 4 (19

REFERENCES1Mudd, S., “A Hypothetical Mechanism of Antibody Production,” J. Immunol.

23, 423 (1932).2Pauling, L., “A Theory of the Structure and the Formation of Antibodies,” J.

Am. Chem. Soc. 62, 2643–2657 (1940).3Dickey, F. H., “The Preparation of Specific Adsorbents,” Proc. Nat. Acad. Sci.

U.S. 35, 227 (1949).4Dickey, F. H., “Specific Adsorption,” J. Phys. Chem. 59, 695 (1955).5Wulff, G., and Sarhan, S., “Use of Polymers with Enzyme-Analogous

Structures for the Resolution of Racemates,” Angew. Chem. Int. Ed. 11, 341(1972).

6Shea, K. J., and Thompson, E. A., “Template Synthesis of Macromolecules.Selective Functionalization of an Organic Polymer,” J. Org. Chem. 43, 4253(1978).

7Shea, K. J., Thompson, E. A., Pandey, S. D., and Beauchamp, P. S., “TemplateSynthesis of Macromolecules. Synthesis and Chemistry of FunctionalizedMacroporous Poly(divinylbenzene),” J. Am. Chem. Soc. 102, 3149 (1980).

8Damen, J., and Neckers, D. C., “Stereoselective Syntheses Via a Photochemi-cal Template Effect,” J. Am. Chem. Soc. 102, 3265 (1980).

9Damen, J., and Neckers, D. C., “Memory of Synthesized Vinyl Polymers forTheir Origins,” J. Org. Chem. 45, 1382 (l980).

10Damen, J., and Neckers, D. C., “On the Memory of Synthesized VinylPolymers for Their Origins,” Tetrahedron Lett. 21, 1913 (1980).

11Zeng, X., and Murray, G. M., “Synthesis and Characterization of Site SelectiveIon Exchange Resins Templated for Lead(II) Ion,” Sep. Sci. Technol. 31, 2403(1996).

12Jenkins, A. L., Murray, G. M., and Uy, O. M., “A Polymer-Based OpticalSensor for the Nerve Agents Sarin and Soman,” Proc. Scientific Conf. onChemical and Biological Research, U.S. Army Edgewood Research, Developmentand Engineering Center (1996).

13Zeng, X., Bzhelyansky, A., Uy, O. M., and Murray, G. M., “A Metal IonTemplated Polymeric Sensor for Lead,” Proc. Scientific Conf. on Chemical andBiological Research, U.S. Army Edgewood Research, Development and Engi-neering Center (1996).

14Sillen, L. G., and Martell, A. E., Stability Constants of Metal–Ion Complexes,The Chemical Society, London, England, 2nd Ed. (1964).

15Yatsimirskii, K. B., and Vasiliev IV, P., Instability Constants of ComplexCompounds, Plenum, New York (1960).

16Shea, K. J., and Stoddard, G. J., “Chemoselective Targeting of FluorescenceProbes in Polymer Networks: Detection of Heterogeneous Domains inStyrene–Divinylbenzene Copolymers,” Macromolecules 24, 1207 (1991).

17Wulff, G., and Akela, A., “Enzyme-Analog Built Polymers, 6. Synthesis of 5-Vinylsalicylaldehyde and a Simplified Synthesis of Some Divinyl Derivatives,”Makromol. Chem. 179, 2647–2651 (1978).

18Jenkins, A. L., Uy, O. M., and Murray, G. M., “Polymer Based LanthanideLuminescent Sensors for the Detection of Nerve Agents,” Anal. Comm. (inpress).

ACKNOWLEDGMENTS: We would like to acknowledge the support of theIndependent Research and Development committee and thrust area panel mem-bers, especially Vincent L. Pisacane, Harvey W. Ko, and Harold I. Heaton, and theTechnical Services Department’s Technical Initiatives Project Office under thedirection of Harry K. Charles, Jr., and Dale W. Wilson.

THE AUTHORS

GEORGE M. MURRAY is a Senior Professional Staff chemist in APL’sTechnical Services Department and Assistant Professor of Chemistry at theUniversity of Maryland Baltimore County. He received a B.A. in chemistry in1982 and a Ph.D. in chemistry in 1988, both from the University of Tennessee,Knoxville. He performed his postdoctoral research at the Transuranium ResearchLaboratory located at the Oak Ridge National Laboratory. Dr. Murray’s researchinterests are centered around developing methods of analysis for the ultratracedetermination of toxic substances in real samples. His laboratory specializes inthe production of imprinted polymers and spectrochemical analysis. He is amember of the American Chemical Society and the Society for AppliedSpectroscopy. His e-mail address is murray@umbc.edu.

97) 471

G. M. MURRAY ET AL.

AMANDA L. JENKINS received a B.S. degree in chemistry from Salisbury StateUniversity in 1993 and an M.S. degree in chemistry from the University ofMaryland Baltimore County in 1995. She is currently a Ph.D. candidate andworks part-time in the chemical analysis laboratory of APL’s MechanicalServices Group. Ms. Jenkins’ current projects include designing polymer-basedfiber-optic sensors. She is a member of the American Chemical Society and theSociety for Applied Spectroscopy. Her e-mail address is ajenki1@gl.umbc.edu.

O. MANUEL UY is a member of APL’s Principal Professional Staff. He has beenengaged in research in space environmental effects, space science instrumenta-tion, reliability of electronic packages, and advanced sensors for environmentalpollution monitoring. He obtained his Ph.D. in physical chemistry from Case-Western Reserve University and was a postdoctoral fellow at Rice University andthe Free University of Brussels, where he specialized in mass spectrometry andhigh-temperature chemistry. Previously, Dr. Uy was a research scientist for rocketpropulsion systems at Space Sciences, Inc., and was a research chemist for arcsand plasmas at General Electric Company. He has published over 100 papers andhas several patents. His e-mail address is Manny.Uy@jhuapl.edu.

ANTON BZHELYANSKY completed three years at the Ivan Pavlov College ofMedicine in St. Petersburg, Russia. In 1995, he received a B.S. degree inbiochemistry and molecular biology from the University of Maryland BaltimoreCounty. He is currently working on his M.S. degree in chemistry from the sameinstitution. Since joining the research group of George M. Murray in 1995, Mr.Bzhelyansky has been involved in the construction of metal ion sensors (notably,for lead and aluminum) based on the ion templated polymer technology. Hisprimary interest concerns the development of fiber-optic sensing devices forenvironmental and biological applications. His e-mail address is abzhel1@gl.umbc.edu.

472 JOHNS HOPKINS APL TECHNICAL DIGEST, VOLUME 18, NUMBER 4 (1997)