PHOTOPOLYMERIZABLE "ROUNDUP" SYNTHESIS, HERBICIDAL ACTIVITY AND COATING FORMULATION

Victoria Piunova

A Thesis

Submitted to the Graduate College of Bowling Green State University in partial fulfillment of the requirements for the degree of

MASTER OF SCIENCE

August 2006

Committee:

Douglas C. Neckers, Advisor

Tomas H. Kinstle

Neocles B. Leontis

iii

ABSTRACT

Douglas Neckers, Advisor

Marine biofouling is a worldwide problem for all seagoing vessels. It causes a roughness of a ship's hull, a decrease in its speed and maneuverability and thereby increasing fuel consumption and emission of waste products into the atmosphere. As long as 2000 years ago, people attempted to prevent biofouling by covering ship’s hulls with copper and lead sheets. Since then a large variety of methods have been tried but none proved ideal. The current research project tests incorporation a glyphosate-based biocide into a model marine coating to prevent the formation of biofilms- one of the first steps in marine fouling and thereby block biofouling process.

This work describes a synthetic route for a novel compound - acrylated glyphosate - and characterization by chemical, analytical and physical methods. Polymerization, photopolymerization and copolymerization experiments proved the novel compound efficiently polymerizes and copolymerizes during reasonably short time intervals

(120sec-10 min). Biological assays based on the Kirby-Bauer test and monitoring of growth inhibition showed that the acrylated glyphosate derivative, as well as its polymer, possess strong herbicidal activity against model and common biofouling organisms.

Incorporation of acrylated glyphosate into model acrylic resin yielded a highly cross- linked coating which proved to be toxic toward the microorganisms. Release experiments showed no leaching of copolymerized acrylated glyphosate from the coating over 21 days. This indicates that the compound, incorporated into the backbone structure of the iv

coating, retains herbicidal activity against common fouling organisms. Therefore, acrylated glyphosate is a promising component for antifouling coatings for seagoing vessels. v

To my beloved parents vi

ACKNOWLEDGMENTS

I would like first express my gratitude to my advisor Dr. Douglas Neckers for his guidance, encouragement and understanding. I would like to thank Dr. Tomas Kinstle and Dr. Neocles Leontis for being on my thesis committee. My special thanks to Dr.

Andrei Federov for valuable advises and discussions. Also, I would like to acknowledge

Dr. Aneta Bogdanova, Dr. Dan Berger, Dr. Bullerjahn, Maria Baranova and my lab mates for they support and help.

vii

TABLE OF CONTENTS

Page

CHAPTER I. INTRODUCTION AND BACKGROUND INFORMATION ...... 1

CHAPTER II. MATERIALS AND METHODS...... 7

Materials ...... 7

General procedure...... 7

Polymerization experiments and polymer characterization...... 7

Thermal polymerization ...... 7

Photopolymerization...... 8

Photo-copolimerization experiment...... 8

Molecular mass determination for poly(acrylated glyphosate) ...... 8

Formulation and characterization of model acylic matrixes loaded with acrylated

glyphosate or glyphosate ...... 9

Release experiment ...... 9

Coating formulation...... 10

Coating characterization ...... 10

Solid state NMR ...... 10

MALDI-TOF Mass Spectrometry ...... 10

SEM ...... 11

MALDI-TOF Mass Spectrometry ...... 11

Synthesis of N-Methacryloyl-N-(phosphonomethyl)glycine (MA) ...... 11

Synthesis of N-Acryloyl-N-(phosphonomethyl)glycine (AA) ...... 12

Computational methods ...... 13 viii

Biological activity tests ...... 14

Media Preparation and Cell Growth Measurements ...... 14

Agar experiments ...... 17

E. Coli Agar experiments ...... 17

Kirby-Bauer test ...... 18

Biological Activity of Coating Pellets ...... 18

CHAPTER III. SYNTHESIS AND CHARACTERIZATION OF ACRYLIC AND

METHACRYLIC DERIVATIVES OF GLYPHOSATE...... 20

CHAPTER IV. POLYMERIZATION OF ACRYLATED AND METHACRYLATED

DERIVATIVES OF GLYPHOSATE...... 31

Molecular mass determination by SEC ...... 39

CHAPTER V. COMPUTATIONAL STUDIES OF ACRYLATED AND

METHACRYLATED DERIVATIVES OF GLYPHOSATE ...... 41

CHAPTER VI. BIOLOGICAL ACTIVITY OF ACRYLATED GLYPHOSATE AND

POLY(ACRYLATED GLYPHOSATE) ...... 44

CHAPTER VII. COATING FORMULATION, CHARACTERIZATION AND

ELUCIDATION OF HERBICIDAL ACTIVITY ...... 53

CONCLUSIONS ...... 61

REFERENCES ...... 63 ix

LIST OF FIGURES

Figure Page

1.1 Sequential steps in biofouling ...... 2

3.1 1H NMR spectrum of N-methacryloyl-N-(phosphonomethyl)glycine ...... 24

3.2 13C NMR spectra of N-methacryloyl-N-(phosphonomethyl)glycine ...... 25

3.3 Two different isomers of N-methacryloyl-N-(phosphonomethyl)glycine ...... 26

3.4 1H NMR spectra of N-acryloyl-N-(phosphonomethyl)glycine ...... 27

3.5 13C NMR spectra of N-acryloyl-N-(phosphonomethyl)glycine ...... 27

3.6 Temperature-dependent 1H NMR spectra of N-methacryloyl-N-

phosphonomethyl)glycine...... 29

3.7 Temperature-dependent 1H NMR spectra N-acryloyl-N-

(phosphonomethyl)glycine ...... 30

4.1 IR spectra of acrylated glyphosate and poly(acrylated glyphosate) ...... 32

4.2 Photopolymerization of AA in the presence of Darocure 1173...... 33

4.3 Photopolymerization of AA in the presence of Irgacure 651 ...... 33

4.4 Polymerization profiles for N-acryloyl-N-(phosphonomethyl)glycine with

different initiators: VA 057, VA 085, VA 086 ...... 35

4.5 Absorption spectra of camphorquinone in 1-hexanol ...... 36

4.6 Double bond conversion versus time in polymerization experiment induced by

visible light with CQ/EDAB as an initiating system ...... 37

4.7 Copolymerization profile for acrylated glyphosate and 2-hydroxyethylene

acrylate mass determination by SEC ...... 38

4.8 SEC profile of poly(acrylated glyphosate) ...... 39 x

4.9 Calibration curve for molecular mass determination of poly(acrylated glyposate) .. 40

5.1 Results of C(P)-N-C=C Amide bond rotational analysis for both AA and MA ...... 41

5.2 Results of C=C-C=O acryloyl bond rotational analysis for both AA and MA ...... 43

5.3 B3LYP/6-31++G(d,p) optimized structures of the lowest energy isomers of

AA and MA ...... 43

6.1 DH 5α growth curve ...... 45

6.2 Number of colonies on M9 agar plates loaded with glyphosate , acrylated

glyphosate, and pure M9 agar ...... 46

6.3 Growth response of Synechococcus 7002 to the treatment with glyphosate and

acrylated glyphosate ...... 47

6.4 Synechococcus 7002, salt sensitivity test ...... 48

6.5 Herbicidal activity test for the agar medium placed over the layer of AA

homopolymer and control ...... 49

6.6 Typical images of Kirby-Bauer test results ...... 50

7.1 Normalized amount of active component released from the coatings as a function

of testing time ...... 53

7.2 Solid state NMR spectra for pure acrylic formulation and formulations loaded

with acrylated glyphosate or glyphosate ...... 55

7.3 MALDI-ToF-MS chromatogram for model resin loaded with AA (0.3M)...... 56

7.4 Images of resin surfaces before and after treatment with water ...... 58

xi

LIST OF TABLES

Table Page

4.1 Initiators for thermal polymerization ...... 31

4.2 Summary of molar extinction coefficients and efficiency of polymerization for

different initiators ...... 35

6.1 Results of the biological activity screening ...... 48

6.2 Average radii of the inhibition zones measured around the filter paper circles

treated with the designated compound ...... 52

7.1 Average radii of the inhibition zones measured around the polymer pellets

containing the designated compound ...... 60

xii

LIST OF SCEMES

Scheme Page

1.1 EPSP catalyzed reaction ...... 4

1.2 Glyphosate degradation pathway ...... 5

3.1 Route I for methacrylated glyphosate synthesis ...... 21

3.2 Route II for allylated glyphosate synthesis ...... 21

3.3 Route III for synthesis of acrylated and methacrylated derivatives of glyphosate ... 22

3.4 Partial character of double bond in dimethyl formamide ...... 25

1

CHAPTER 1. INTRODUCTION AND BACKGROUND INFORMATION.

Biofouling is a major problem for all seagoing vessels. The US Navy Department alone spends about 1.7 billion dollars annually to overcome this1.

Biofouling is usually defined as an undesirable attachment of microorganisms (i. e. algae, barnacles, mussels) to a surface in contact with fresh or marine water for a period of time2. The crust of marine organisms rapidly formed on a ship’s hull significantly decreases the speed and maneuverability of the marine vessel which, in turn, increases the consumption of diesel fuel and the emission of toxic waste products into the atmosphere. Another problem related to biofouling is the transport of marine organisms from one marine ecosystem to another. This may cause ecological problems in some ecosystems by introducing exotics that can grow uncontrollably.

Owing to the reasons cited above, biofouling prevention is a subject of intense research interest. Recently, four sequential steps of marine fouling mechanism were proposed.3, 4 Any body inserted into marine or fresh water immediately starts to accumulate dissolved organic and molecules by adsorption. This is refers to the first step in fouling mechanism. Bacteria and single cell diatoms settle on a surface of the vessel and secrete sticky muco-polysaccharides (natural glue) leading to the formation of the microbial biofilm. This represents the second step in the marine fouling. The presence of adhesive exudates helps to trap more marine microorganisms.

Eventually, larger marine invertebrates (barnacles, limpets, sea mats and sea squirts) attach and grow together with the macroalgae. The steps of marine biofouling are represented schematically in Figure 1. Attempts, as far back as 2000 years ago, were made to protect marine vessels against fouling by marine organisms. Early Carthaginians used copper sheathing on ships’ bottoms.5 Wax, tar and asphalt were used by other ancient cultures. Later, from the 13th to

15th centuries, pitch was extensively used to protect ships’ hulls6. Two centuries later pitch was 2 replaced bycopper and lead sheathing. This method remained successful and widely used until the middle of the 19th century when the first antifouling paints were invented. All paints incorporated a toxic additive (copper oxide, mercury oxide, arsenic) into the paint matrix. Some of those formulations were relatively successful but all of them still possessed short lifetimes and were expensive.

Figure 1.1. Sequential steps in biofouling. Figure adapted from ref. 3

Nevertheless, those paints were in use until 1950’s when tert-butyltin (TBT) was invented by Van de Kerk and co-workers.7 Organotin compounds subsequently became extensively used to prevent marine fouling. Even today coatings with incorporated tert-butyltin are still widely used to prevent biofouling. The main disadvantage of these coatings is their high toxicity, which affects not only algae and barnacles, but also phytoplankton and marine animals.

Because of this, the use of TBT containing paints will be completely banned by 2007. 3

Low surface energy coatings (foul release coatings) are considered a promising

replacement for tin containing coatings. The mode of action is based on the idea that marine

organisms will have difficulties settling on a surface of these coatings. As long the ship is in motion, water shear will remove all organisms settled on the hull’s surface. Silicones and

fluoropolymers are materials of interest for this class of coatings. Experimental results for this

type of coating demonstrated the reduction of fouling coverage of more than 20% in three years.

But at the same time there are many disadvantages: the coatings are expensive, exhibit poor

adhesion properties, can be easily damaged and not efficient in static harbor conditions.8

Therefore, effective, non-expensive foul release coating with good mechanical properties are still to be invented.

The purpose of the present project is to develop and investigate the antifouling properties of a biocide-containing a photopolymerizable acrylic coating. It is worth pointing out that photopolymerization is an attractive technology for coating formulation. In contrast to solvent born coatings, photopolymerizable coatings need no time for drying after application. This gives advantages to photopolimerizable coatings. First, the coating process can be finished within minutes. Second there is no pollution by toxic volatile organic solvent. The acrylic derivative of glyphosate was chosen as the biocide. The choice to use glyphosate as a biocide was made based on factors.

Glyphosate or Roundup® is a well-known non-selective systematic herbicide. It is an

effective inhibitor of the shikimic acid pathway in microorganisms, weeds, trees and bush

species. The target of glyphosate action is 5-enolpyruvylshikimate 3-phosphate (EPSP) synthase,

an enzyme essential for the synthesis of aromatic amino acids9, 10. EPSP synthase catalyzes the

transfer of the enolpyruvyl moiety from phosphoenol pyruvate (PEP) to shikimate-3-phosphate 4

(S3P) forming the EPSP and inorganic phosphate as products (Scheme 1). Glyphosate inhibits

EPSP synthase in a slow reversible reaction, which is competitive versus PEP and uncompetitive versus S3P11.

Scheme 1.1. EPSP catalyzed reaction. Scheme adapted from ref. 11

Because EPSP synthase does not exist in mammals and other marine animals, glyphosate would be an attractive biocide for antifouling applications. Another attractive feature of glyphosate is the possibility of its degradation by some microorganisms. For example, soil bacteria Geobacillus caldoxylosilyticus efficiently degrades glyphosate in two weeks according

to pathway represented in scheme 1.2.12 Therefore, it does not persist in the environment.

5

Scheme 1.2. Glyphosate degradation pathway. Scheme adapted from ref. 12

We anticipate several advantages from incorporation of glyphosate as a biocide into a

coating. First, glyphosate will only affect target organisms in the marine environment (i. e. algae, diatoms, bacteria), preventing biofilm formation on the surface of interest and, as a consequence,

other biofouling steps. Another important advantage is the degradation of glyphosate by some

types of microorganisms. The glyphosate leached from the coating will be degraded by microorganisms and will not accumulate in the ocean. On the other side, it is unlikely that those

organisms will be capable of degrading glyphosate in the coating. Owing to their physiology

they are unable to stick to the surface (coated surface in particular).

The objectives of this work are (a) to functionalize glyphosate with the acrylic moiety to

create a monomer that can be incorporated into photopolymerizable acrylic coating (b) to

investigate the radical polymerization behavior of this monomer using different photo- and 6 thermal initiators, as well as different light sources, (c) to test the herbicidal activity of functionalized glyphosate and compare it with that of an unfunctionalized glyphosate (d) to incorporate functionalized glyphosate into the model coating formulations, and (e) to characterize the novel formulations and investigate their activity against common marine fouling organisms.

7

CHAPTER II. MATERIALS AND METHODS.

Materials

N-(phosphonomethyl)glycine was obtained from Aldrich and used as received. Acryloyl

chloride (Aldrich) and methacryloyl chloride (Alfa Aesar) were distilled prior to use. Darocur®

® 1173 and Irgacure 651 were obtained from Ciba Specialty Chemicals, Inc., tert-butyl peroxide

was obtained from Aldrich and was used as received. The water-soluble diazo initiators, VA-

086, VA-085, and VA-057 were gifts from Wako Pure Chemical Industries, Ltd. The visible

light photoinitiation system containing camphorquinone (CQ) (Aldrich) and ethyl –p-(dimethyl

amino benzoate (EDAB) (Aldrich) was used as received.

General procedure

Nuclear magnetic resonance (NMR) spectra were recorded on Varian Unity + 400 MHz

or Bruker Avance 300 MHz spectrometers. The NMR experiments at elevated temperatures were

carried out on a Varian Unity + 400 MHz spectrometer. All NMR spectra were recorded in

deuterium oxide (D2O), chemical shifts are reported in ppm and referenced to 3-

(trimethylsilyl)propionic acid-d4, sodium salt (TSP). FTIR spectra were recorded on a

ThermoNicolet IR200 spectrometer. High- resolution mass spectral analyses were performed by

the mass spectrometry laboratory, University of Illinois at Urbana-Champaign, IL.

Polymerization experiments and polymer characterization

I. Thermal polymerization. The acrylic acid derivative of glyphosate (2 mmol, 130 mg)

was dissolved in 2 ml of water. The initiator, 1% wt (VA-057, VA-085 or tert-butyl peroxide)

was added and the reaction mixture was heated to the decomposition temperature of initiator, for

a half-life period of the initiator. Polymers were isolated by precipitation with 2ml of ethyl

alcohol. The products of polymerization were characterized by IR- and mass-spectrometry. 8

II Photopolymerization. Polymerizations were carried out by irradiation of ca. 17%

® solutions of glyphosate derivatives in D2O in the presence of 1.5 wt % initiator (Darocure 1173 ) using a RMR-600 Rayonet photochemical reactor equipped with seven lamps (λex=350nm).

Visible light polymerization experiment was done by irradiation of a 20% solution of acrylated

glyphosate in D2O containing 2% of CQ and 2% (EDAB) using Xenon lamp with broad band

filter (λ = 400 – 600 nm) as a light source. The double bond conversion of a monomer was followed with 1H NMR using the TSP signal as internal reference.

III Photo-copolymerization experiment Photo-copolymerization was carried out by

irradiation of ca. 17% solutions of glyphosate derivative and 17% of 2-hydroxy-ethyl acrylate

® monomer in D2O in presence of 1.5 wt.% (Darocure 1173 ) using a RMR-600 Rayonet

photochemical reactor equipped with seven lamps (λex=350nm). Monomer double bond

conversion was monitored by 1HNMR.

Molecular mass determination for poly(acrylated glyphosate).The molecular mass of

homopolymer of acrylated glyphosate was determined by h.p.l.c. size-exclusion column

chromatography (Bil-Sil SEC-300; Bio-Rad; 300 mm x 7.5 mm). Chromatography was carried

out with water/acetonitrile (9 : 1) mobile phase at a flow rate of 1 ml/min. 9

Formulation and characterization of model acylic matrixes loaded with

acrylated glyphosate or glyphosate.

Copolymer pellets (d=5mm, h=2mm) were formed from a model acrylic resin

(designated as 459S2) containing several epoxy acrylic oligomers (57%), acrylic monomeric

diluents (43%) and a corresponding amount of herbicide (glyphosate or acrylated glyphosate). In

1g of liquid resin 459S2, 2.4, 3.2, or 6% wt. (0.1, 0.5 or 0.3 M, respectively) of acrylated

glyphosate or 1.6, 2.4, or 5% wt. (0.1, 0.15 or 0.3 M, respectively) of glyphosate were dispersed

or partially dissolved by using ultrasonication and oven heating (80°C, 5 min). To each sample,

1.5% wt. of commercial photoinitiator (Irgacure 819) was added. Liquid formulations were

poured into an open top plastic mold and irradiated at 350 nm under inert atmosphere (Ar of N2)

for 2 min. After completing the irradiation, copolymer pellets were removed from the form.

Pellets of the three listed glyphosate concentrations were used for the biological activity tests

(wide infra). Only pellets containing 0.3 M of active ingredient were used for release experiments.

Release experiments. It was experimentally established that glyphosate and acrylated

glyphosate have equally good solubility in water. Therefore, it was possible to monitor the

amount of both acrylated and unfunctionalized glyphosates released from the coating into an

aqueous media. Synthesized pellets (with concentration of herbicides of 0.3M) were emerged

into 1ml of D2O and kept on a shaker (100 rpm) for 20 days. The amounts of released herbicide

were periodically monitored by 1H NMR analysis of pellet washings through observing the intensities of characteristic peaks of acrylated glyphosate (3.9 d ppm and 4.4 d ppm) and unfunctionalized glyphosate (3.2 and 3.9 ppm). The concentration of herbicide released was normalized relative to the NMR signals obtained from the standard D2O solutions of glyphosate 10

and acrylated glyphosate with a concentration equal to that loaded into the pellets of the

corresponding herbicide.

Coating formulation. Copolymer coatings were made using the same model acrylic resins

loaded with the same amount of glyphosate and acrylated glyphosate as for pellets formulation.

In addition, they contained 2% of isopropylthioxanthone (ITX) (Albemarle®) photoinitiator and

2% of octyl-p-dimethylamino benzoate (ODAB) (Aldrich) synergist. Plane microscope slides

(3”x1”x1mm) were covered with 0.01g of liquid formulations and irradiated under inert

atmosphere (N2) at 395 nm using Clearstone Tech UV-Vis LED source (CF-1000).

Coating characterization

Copolymer pellets with 0.3M glyphosate and 0.3M acrylated glyphosate content were

crushed into fine powders and used for qualitative analysis of the coating. Glass slides coated

with the resins containing the same amounts of glyphosate and acrylated glyphosate were also

used.

Solid state NMR. Solid state NMR spectra were recorded on a Varian Unity + 400 MHz spectrometer. All spectra acquired using 5mm rotor and a spinning speed of 5000 rpm.

Chemical shifts were recorded in ppm.

MALDI-TOF Mass Spectrometry. MALDI-TOF-MS analysis was carried out on a

Bruker OMNIFLEX MALDI mass-spectrometer. Trans-2-[3-(4-tret-butylphenyl)-2-methyl-2-

propenylidene] malononitrile (DCTB) obtained from Aldrich (HPLC pure) was used as the

matrix for analysis. The matrix was dissolved in THF at a concentration of 40 mg ml-1.

Potassium trifluoroacetate (Aldrich 98%) was used as the ionization agent and was added to THF

at a concentration of 1 mg ml-1. The copolymer was dissolved in THF in the amount of 2 mg ml- 11

1. In a typical MALDI-TOF-MS experiment, the matrix, salt and polymer solutions were

premixed in a ratio of 10:1:5.

SEM. The scanning electron microscopy of the coating surface was obtained on a

“Hitachi S- 2700 SEM” at accelerating voltage of 15 kV with a 100- and 350-fold magnification.

Images of freshly prepared coatings and those of coating surfaces after 13 days of water

treatment were investigated by SEM and compared to observe structural changes.

Sample Preparation for SEM Analysis.

Coated glass slides were cut into smaller pieces and were placed firm onto aluminum stubs using

graphite glue. The glue was allowed to air dry. In order to enhance contrast of the images, slides

were sputter coated with a 6 nm thick gold/palladium film.

Synthesis of N-Methacryloyl-N-(phosphonomethyl)glycine (MA)1

The synthesis of MA was carried out according to the method reported by Gough13 with several modifications. N-(phosphonomethyl)glycine (0.169g, 1mmol) was dissolved in 2.4 mL

10% sodium hydroxide solution. Then, to a vigorously stirred solution, methacryloyl chloride

(0.146 mL, 1.5mmol) was added dropwise under argon. After continuous stirring for four hours at room temperature, the reaction mixture was acidified with hydrochloric acid to pH=0 or below, and concentrated under vacuum to precipitate the NaCl, which was removed by filtration.

The pH of the filtrate was then adjusted to 1-2 by adding sodium hydroxide solution. Further concentration of the reaction mixture yielded white precipitate, which was filtered and dried

1 under vacuum to afford MA as a white solid (0.189g, 40%). H NMR (400 MHz, D2O, δ/ppm):

2 2 1.90 (s, 3H, CH3), 1.95 (s, 3H, CH3), 3.82 (d, JHP = 12 Hz, 2H, CH2-P), 3.87 (d, JHP = 12 Hz,

2H, CH2-P), 4.32 (s, 2H, CH2), 4.48 (s, 2H, CH2), 5.12 (s, 1H, HC=C), 5.29 (s, 1H, HC=C ), 5.32

1 This work was largely done by Dr. A. Bogdanova 12

13 (s, 1H HC=C), 5.41 (s, 1H, HC=C ). C NMR (75 MHz, D2O, δ/ppm): 19.52 (CH3), 19.63

1 1 (CH3), 43.44 (d, JPC =147 MHz, -CH2-P), 47.98(d, JPC=147 MHz, -CH2-P ), 48.54 (-CH2-C=O),

51.64(-CH2-C=O), 117.63 (CH2=C-), 118.30 (CH2=C-), 138.82 (CH2=C-),139.28 (CH2=C-),

-1 172.64 (CH2-C=O), 173.30 , (CH2-C=O) 175.90, (N-C=O) 176.45 (N-C=O). IR (neat, cm ):

+ + 1720, 1678, 1590, 1462, 1406. HRMS (ES ): Calcd. for C7H13NO6P [M+H] : 238.0481. Found:

238.0487.

Synthesis of N-Acryloyl-N-(phosphonomethyl)glycine (AA)2

AA was synthesized using a similar procedure to that described for MA. Acryloyl

chloride was used in one-third-fold excess relative to [glyphosate]. The reaction was completed

1 in one hour to give AA as a white solid in 50% yield. H NMR (300 MHz, D2O, δ/ppm): 3.95 (d,

2 2 JHP = 12 Hz, 2H, CH2-P), 3.98 (d, JHP = 12 Hz, 2H, CH2-P), 4.29 (s, 2H, CH2), 4.50 (s, 2H,

CH2), 5.87 (dd, AMX system, JAX = 10.8 Hz, JAM = 1.2 Hz, 1H, H2 C=CH), 5.93 (dd, AMX

system, JAX = 10.8 Hz, JAM = 1.2 Hz, 1H, H2 C=CH), 6.25 (dd, AMX system, JMX = 16.8 Hz, JMA

= 1.2 Hz, 1H, H2C=CH), 6.29 (dd, AMX system, JMX = 16.8 Hz, JMA = 1.2 Hz, 1H, H2C=CH),

6.61 (dd, AMX system, JXM = 16.8 Hz, JXA = 10.8 Hz, 1H, H2C=CH), 6.82 (dd, AMX system,

13 JXM = 16.8 Hz, JXA = 10.8 Hz, 1H, H2C=CH). C NMR (75 MHz, D2O, δ/ppm): 44.24 (d,

1 1 JPC=150 MHz, -CH2-P), 46.73 (d, JPC=150 MHz, -CH2-P), 50.04 (-CH2-C=O), 50.74 (-CH2-

C=O), 126.86 (CH2=C-), 126.96 (CH2=C-), 130.27 (CH2=C-), 130.65 (CH2=C-), 169.94 (CH2-

-1 C=O), 169.96 (CH2-C=O), 172.66 (N-C=O), 172.75 (N-C=O). IR (neat, cm ): 1722, 1637, 1571,

+ + 1472, 1401. HRMS (ES ): Calcd. for C6H11NO6P [M+H] : 224.0324. Found: 224.0330. Calcd.

+ for C6H10NO6PNa [M+Na] : 246.0140. Found: 246.0142.

2 This work was largely done by Dr. A. Bogdanova 13

Computational methods3

Computational studies were performed for N-acryloyl-N-phosphonomethylglycine (AA) and N-methacryloyl-N-phosphonomethylglycine (MA) using Spartan’06 (ref) except for

B3LYP/6-31++G(d,p)// B3LYP/6-31++G(d,p) calculations which were performed using

Gaussian’03.14

PM3 semi-empirical calculations and density functional methods with diffuse basis sets were

used in order to conserve their hydrogen-bonded character during optimization.

O O O O a b a b HO P CH2 NCH2 COH HO P CH2 NCH2 COH HO c HO c d C d C OC3 2 H OC3 2 CH3 4 4 1 CH2 1 CH2

AA MA

Structures were selected to maximize hydrogen bonding stabilization. Some structures were

chosen in which the phosphoryl group was a hydrogen-bond donor to the acryloyl carbonyl,

while the carboxyl group was a hydrogen-bond donor to the negative phosphoryl oxygen. Other

structures were chosen with the carboxyl group as a hydrogen-bond donor to the acryloyl

carbonyl, while the phosphoryl group was a hydrogen-bond donor to the carboxyl group.

Rotational studies were performed, in which one of two chosen dihedral angles was constrained to a series of values and the rest of the molecule was allowed to relax. Several series of structures were generated by constraining a certain dihedral angle—either the acryloyl C=C-

C=O dihedral (1-2-3-4) or the amide C(P)-N-C=O dihedral (a-b-c-d) and allowing the rest of the

structure to relax by sequential molecular mechanics, PM3 and B3LYP/6-31G calculations.

Geometries at and near the local minima identified by the rotational studies were fully optimized

(no dihedral angle constraints) at the B3LYP/6-31+G* level.

3 This work was largely done by Dr. D. Berger and Dr. A. Fedorov 14

Eight candidate local minima were found for MA, with vinyl dihedral angles of ±120-

130º and ±40-60º. Amide dihedral angles were all within 15º of planarity. Nine candidate local minima were found for AA, with vinyl dihedral angles from -6º to +8º; from +127º to +132º; and

from -126º to -132º. (see Supporting Information). Candidate local minima were fully optimized

at the B3LYP/6-31++G(d,p) level. The absence of imaginary frequencies from vibrational

analysis confirmed that all structures found were indeed true local minima.

Biological activity tests

The following materials were tested for biological activity: glyphosate (Monsanto

Chemical Co.), acrylated glyphosate (AA), and poly(acrylated glyphosate). Glyphosate and

acrylated glyphosate were dissolved in doubly deionized water (200 mM), (the reported concentration is adjusted for the maximum possible content of the residual NaCl in the acrylated gylphosate) and these stock solutions were used to prepare all relevant biological media.

Polymerization of acrylated glyphosate was also conducted in aqueous media (350 nm irradiation with 1.5% wt. VA-085 initiator), with the same initial concentration of the monomer as for the

unconverted AA experiments.

Strains of green algae, cyanobacteria and E Coli were used as test organisms. The green

algae CD 1 Red and cyanobacteria Synechocystos 6803 were obtained from the collection of

Prof. George Bullerjahn. Salt water strain of cyanobacteria Synechococcus 7002 was obtained

from the collection of Prof. Michael McKay. E Coli DH 5α was obtained from Invitrogen Inc,

diatoms Phaeodactylum tricornutum and Navicula pelliculosa were obtained from National

Center for Culture of Marine Phytoplankton.

Media Preparation and Cell Growth Measurements. Green algae and cyanobacteria were

propagated photoautotrophically in 250 mL Erlenmeyer flasks containing 100 mL of liquid BG- 15

11 culture medium (for 1L: NaNO3 10ml (15g/100ml); CaCl2x2H2O 1ml (3.6g/100ml);

FeNH4Citrate 1ml (1.2g/100ml); Na2EDTA 1ml (0.1g/100ml); K2HPO4 1ml (4.0g/100ml);

MgSO4 1ml (7.5g/100ml); Na2CO3 1ml (2.0/100ml); Micronutrients 1ml – Contain: H3BO3

2.8g/L, MnCl2x4H2O 1.8g/L, ZnSO4x7H2O 0.222g/L, Na2MoO4x2H2O 0.39g/L, CuSO4x5H2O

0.079g/L, CoCl2x6H2O 0.012g/L. The grown culture medium was sterilized in an autoclave at

121 oC and 1.05 kg cm-2 pressure for 20 min. For the cell experiments, 20 mL aliquots of the BG-

7 -1 11 medium containing the green algal ([cells]initial = 5.9 x 10 mL ) or cyanobacteria cells

9 -1 ([cells]initial = 1.7 x 10 mL ) were placed in sterile 50 mL Erlenmeyer flasks. Cell concentration was determined by counting with hemacytometer (Hemacytometer Fisher scientific) under light microscope.

The cultures were kept on a rotating shaker (100 rpm) at 25oC for 5 days and

continuously illuminated with cool-white fluorescent light with a constant light intensity of 5000

lux. Two types of diatoms: Phaeodactylum tricornutum and Navicula pelliculosa were propagated autotrophycally in 250 ml Erlenmeyer flask containing 2ml of dense culture with 100 mL of liquid f/2 culture medium (For 1L of f/2 medium: 1ml NaNO3 (75g/L dH2O), 1ml

NaH2PO4xH2O (5g/L dH2O), 1ml Na2SiO3x9H2O (30 g/L dH2O), 1ml f/2 trace metal solution

(for 1L of f/2 trace metal solution: 3.15 g FeCl3x6H2O, 4.36 g Na2EDTAx2H2O, 1ml

CuSO4x5H2O (9.8 g/L dH2O), 1ml Na2MoO4x2H2O (6.3 g/L dH2O), 1ml ZnSO4x7H2O (22.0

g/L dH2O), 1ml CoCl2x6H2O (CoCl2 · 6H2O), 1ml MnCl2x4H2O (180.0 g/L dH2O) and ddH2O to total volume of 1L), 0.5 ml f/2 vitamin solution (for 1L of vitamin solution: 1ml cyanocobalamin

(1.0g/L dH2O), 10 ml biotin (0.1g/1L dH2O), 200mg thiamine). The grown culture medium was

sterilized in an autoclave at 121 oC and 1.05 kg cm-2 pressure for 20 min. Cultures were kept on a 16

rotating shaker (100 rpm) at 25oC for 10 days and continuously illuminated with cool-white

fluorescent light with a constant light intensity of 5000 lux.

CD 1 Red, Synechocystos 6803 and Synechococcus 7002 were then treated with an

aliquot of a stock herbicide solution to bring the final herbicide concentration to 1 mM and

incubated for 7 days on a rotator shaker (100 rpm) at 20 oC and continuous light intensity of

5000 lux. Cell counts were correlated with the absorbance and measured as a function of time using a Spectronic 20 Genesis spectrophotometer. As previously reported by Ma et al.15, the optimal wavelength for monitoring culture growth is 680 nm. Experiments for each herbicide were repeated three times. Control and treated cultures were grown under the same conditions of temperature, photoperiod, and agitation. In each experiment, the inhibition of cell growth in treated cultures was monitored spectrophotometrically and compared to the growth in control samples.

E. Coli was propagated heterotrophically in a 2 mL Eppendorf test-tube containing 1.5

mL of liquid M9 minimum medium (for 1 L: 20 ml M9 salt (per 1 L: Na2HPO4 33.9g, KH2PO4

15g, NaCl 2.5g, NH4Cl 5g), 1M MgSO4 2ml; 20% glucose solution 20ml; 1M CaCl2 0.1ml) and

kept on a rotating shaker (100 rpm) at 37oC for 12 hours. The grown culture medium was

sterilized at 121oC, 1.05 kg cm-2 for 20 min. For the cell experiments, 100 mL aliquots of the M9

5 -1 medium containing the E. Coli cells ([cells]initial = 6 x 10 mL ) were distributed into sterile

250mL Erlenmeyer flasks and treated with a stock herbicide solution to obtain a final concentration of 30mM, followed by an incubation for 24 h on a rotating shaker (100 rpm) at

37oC. Cell counts were correlated with the absorbance at 600 nm over time using a Spectronic 20

Genesis spectrophotometer. Control and treated cultures were grown under the same conditions 17

of temperature and shaking. In each experiment, the inhibition of growth in treated cultures was

monitored spectrophotometrically and compared to the growth in control samples.

Agar experiments. BG-11 medium containing 1.5 wt. % agar was prepared and sterilized

at 121oC, 1.05 kg cm-2 for 20 min. Each herbicide tested was prepared as a 30 mM solution and

added to the autoclaved agar medium to yield a 1 mM final concentration of the herbicide. A 20

mL aliquot of the liquid agar was transferred into a standard Petri dish (100x15 mm) and allowed to harden at room temperature. Cell suspensions of green algae CD 1 Red, cyanobacteria

Synechocystos 6803 and Synechococcus 7002 were inoculated using a sterile inoculation loop.

Dishes were stored at 20oC under illumination with the cool-white fluorescent light having a

constant light intensity of 5000 lx. The experiment for each herbicide was repeated five times.

Control and treated cultures grew under the same conditions of temperature and photoperiod. In

each experiment, the inhibition of growth in treated cultures was monitored relative to the

growth in control samples using qualitative visual observation. Additional experiments for

testing the toxicity of the polymer were implemented. Ten mL of aqueous polymer solution was

placed on a sterile Petri dish and left at room temperature until all water was evaporated (3 days)

to form a thin polymer film. Twenty mL of the liquid agar was transferred into the Petri dish,

poured on a polymer film and allowed to harden at room temperature. Cell suspensions of green

algae CD 1 Red, cyanobacteria Synechocystos 6803 and Synechococcus 7002 were inoculated

using a sterile inoculation loop. An analogous system containing no polymer film was used as a

control for the experiment. Inhibition of growth in treated cultures was monitored relative to the

growth in control samples using qualitative visual observations.

E. Coli Agar experiments. A M9 minimum medium containing 1.5 wt. % agar was

prepared and sterilized at 121 oC, 1.05 kg cm-2 for 20 min. An aliquot of each tested herbicide 18

solution was added to portions of the autoclaved agar to give a 1 mM final concentration of the

herbicide. Twenty mL of the liquid agar was poured into a standard Petri dish and allowed to

harden at room temperature. E. Coli cell suspensions were inoculated using a sterile inoculation loop and stored at 37oC for 24 hours. Control and treated cultures were grown under the same

conditions. In each experiment, the inhibition of growth in treated cultures was monitored

relative to growth in control samples by manually counting the number of colonies.

Kirby-Bauer test. The herbicidal activity of novel acrylated glyphosate and

poly(acrylated glyphosate) as well as acrylamide and glyphosate (as controls) was also compared

using the Kirby-Bauer disc diffusion method.16 Filter paper circles (d = 5 mm) soaked with

acrylated glyphosate, poly(acrylated glyphosate), glyphosate and acrylamide solutions of given concentrations were placed on the Petri plates, loaded with a layer of BG-11 containing agar medium (1.5% wt., 10 ml) and a layer of BG-11 containing low melting agarose medium (0.8% wt., 5 ml) previously seeded with green algae CD 1 red (5.9x106 cell/ml, 2 ml) or cyanobacteria

Synechococcus 7002 (1.7x108 cell/ml, 2 ml). The same type of experiment was introduced for diatoms Phaeodactylum tricornutum and Navicula pelliculosa, but instead BG-11, f/2 medium was used. After phototrophical propagation during the 5 days for cyanobacteria and green algae and 15 days for diatoms under cool fluorescence light (intensity 5000 lux), the average radii of inhibition zones were measured. Water soaked filter paper circles were used as controls.

Experiments for each herbicide concentration were repeated 3 times.

Biological Activity of Coating Pellets. The biological activity of model coating

formulations containing AA copolymerized with other monomers and oligomers was also tested.

Two types of diatoms cultures - Phaeodactylum tricornutum and Navicula pelliculosa were used

in the tests. Polymer pellets loaded with different concentrations (0.1, 0.15, and 0.3 M) of 19 glyphosate (control) and acrylated glyphosate were placed on Petri plates loaded with a layer of f/2- containing agar medium (1.5% wt., 10 ml) and a layer of f/2-containing low melting agarose medium (0.8% wt., 5 ml) previously seeded with diatoms. After phototrophical propagation during 15 days under cool fluorescent light (5000 lx), the average radii of inhibition zones were measured. Experiments for each herbicide concentration were reproduced at least 2 times. Pellets formed of a model acrylic resin 459S2 without glyphosate additives were used as a control. 20

CHAPTER III. SYNTHESIS AND CHARACTERIZATION OF ACRYLIC AND

METHACRYLIC DERIVATIVES OF GLYPHOSATE4.

Two major challenges that arise in the synthesis of glyphosate derivatives are poor solubility in

organic solvents and the presence of both acidic and basic functional groups in the same

molecule. Glyphosate exists as a zwitterion.17 Acid dissociation constants for glyphosate are

pKa1 0.8 (first phosphonic), pKa2 2.3 (carboxylic), pKa3 6.0 (second phosphonic), and pKa4 11

(amine). Further, the carboxyl and phosphonyl groups are electrophilic, while the amino group is

nucleophilic. The double-bonded and deprotonated oxygens of the phosphonyl group are

stabilized by delocalization resulting in the phosphonyl group being less susceptible to a nucleophilic attack than the carboxyl group.

Attempts to utilize standard esterification routes were unsuccessful.

N-(Phosphonomethyl)glycine was converted to the corresponding acyl chloride by reaction with

thionyl chloride. 18 When, followed by treatment with methacrylic acid in the presence of

triethylamine (synthetic route I), a complex mixture resulted. The NMR spectrum of this mixture

showed no olefinic signals. Thus, the reaction products were not separated. The product of

reaction of glyphosate with methacryloyl chloride also did not represent target product as

indicated by the H1NMR spectra.

4 This work was largely done by Dr. A. Bogdanova 21

Scheme 3.1. Route I for methacrylated glyphosate synthesis.

CO2H O H OH O H Cl O H O O SOCl2 OH P N OH P N OH P N O O O OH THF OH Et3N OH

The successful procedure involved activating the glycine carboxylic acid functionality

with chlorotrimethylsilane forming the corresponding silyl ester. This glycine derivative then

reacts with alcohols to generate the desired glyphosate esters that were isolated by treating the

reaction mixture with propylene oxide (synthetic route II).18 Using this procedure, we have

obtained the allyl ester of glyphosate in a quantitative yield. Nevertheless, all other efforts to

prepare more readily photopolymerizable anhydrides or esters, such as those of methacrylic or

acrylic acids, were unsuccessful.

Scheme 3.2. Route II for allylated glyphosate synthesis

Cl Si O H OH 1. O H O OH P N OH P N + OH O O 2. O OH OH 100%

Preparation of the methacrylated amide of a glyphosate was achieved as follows: N-

(phosphonomethyl)glycine was dissolved in a 6-fold excess of 10% NaOH solution and subsequently treated with the corresponding acyl chloride to form the amide derivative,

according to scheme 3.3. 22

Scheme 3.3. Route III for synthesis of acrylated and methacrylated derivatives of glyphosate

In order to obtain a pure product in a reasonable amount at once, we had to overcome two substantial problems. First, because of the 6-fold excess of NaOH required for successful synthesis, substantial amount of NaCl appeared as a side product. Second, the reasonable amount of product was only possible to obtain on a small scale (1-2 mmol of glyphosate).

The purification might be accomplished based on differences of solubility in water of methacrylated glyphosate or acrylated glyphosate and that of NaCl at pH below 0. After reaction

(checked by 1H NMR) was completed, the pH of the mixture was adjusted to below zero by adding a concentrated hydrochloric acid. At pH=0, the solubility of NaCl in H2O is substantially lower than that of the target product and precipitated alkali metal salt was removed by filtration.

The pH of the aqueous solution was then adjusted to 1-2 by adding NaOH solution.

Concentration of the reaction mixture under vacuum led to the precipitation of the desired amide, which was substantially more pure. Similar observations have been made for the solubility of glyphosate and a method for its isolation from aqueous alkali metal solutions has been previously proposed.19 Despite our extensive efforts to reduce the NaCl content in the final product, the residual NaCl amount remains at about 30-40% level. 23

Scaling up of the reaction was done only for acrylated glyphosate because methacrylated

glyphosate failed to produce the polymer (wide infra). However, all considerations described for

acrylated glyphosate are entirely applicable to methacrylated species. The conditions (addition of

reagent, temperature, inert atmosphere) were conserved as well as ratios of reagents: 6-fold

excess of NaOH and 1.3 fold excess of acryloyl chloride. After 3 minutes of reaction, several drops of the reaction mixture were concentrated under vacuum and the resulting white solid was

1 dissolved in D2O and analyzed by H NMR. This allowed us to determine the peak intensities of

starting materials and monitor with time the remaining amounts of starting materials in the

reaction mixture. Next 1H NMR analyses was done every hour with the peak intensities of the

remaining starting materials (glyphosate characteristic peaks at 3.18, 3.23 and 3.98 ppm)

compared to the initial intensities. Based on this information, we were able to calculate and add into reaction mixture the additional amounts of acrylic acid chloride required to convert remaining glyphosate into corresponding acrylamide. This “step- by- step” technique helped to accomplish conversion of 100 mmol of glyphosate into its acrylated analog with a 95% yield. It is worth noting that increasing the concentration of acryloyl chloride in the original reaction mixture was not successful and the product yield was no more that 15%. Another interesting observation was that according to the 1H NMR, all acryloyl chloride was consumed within first

15 minutes of the reaction. This might be because of hydrolysis of acryloyl chloride by

hydroxide ions producing is less reactive acrylic acid, which is not capable of reacting with

glyphosate. The reasonable solution for this issue might be decrease of the amount of the NaOH

in the reaction mixture. But experimental results showed that degree of conversion did not

increase with reduction of NaOH amount. For successful reaction, the glyphosate should be

completely deprotonated. As far as this compound possesses four acidic and basic functionalities, 24

at least 4-fold excess of NaOH is needed. Experimentally it was found that optimum amount

NaOH is in 5-6 fold excess.

The NMR spectra of the methacrylic and acrylic acid derivatives of glyphosate are shown

in Figures 1.3, 2.3, 4.3, 5.3. The 1H NMR spectrum of N-methacryloyl-N-

(phosphonomethyl)glycine (MA) in deuterium oxide (Figure 3.1) shows two identical sets of

signals in a 1:1 ratio for all protons: a doublet for the methyl group, a doublet of doublets for a

methylene group attached to phosphorus, two singlets for a methylene group attached to the

carboxylic group, and two different double bond signals. Similar situation was observed in 13C

NMR spectra, where all signals are doubled (Figure 3.2). We believe the reason for this phenomenon is an existence of two equally stable isomers of methacrylated glyphosate.

Figure 3.1. 1H NMR spectrum of N-methacryloyl-N-(phosphonomethyl)glycine (MA)

25

Figure 3.2. 13C NMR spectra of N-methacryloyl-N-(phosphonomethyl)glycine (MA).

It was established previously that amide exists in two different conformations due to hindered rotation around the C-N bond, which possesses a partial double bond character.20 Because the

rate of rotation is slow, it is possible to observe two different species. The classical example of

this equilibrium is dimethylformamide, where the 1H NMR spectrum shows two resolved signals

for the methyl groups (Scheme 3.4).

Scheme 3.4. Partial character of double bond in dimethyl formamide

O CH3 O CH3 + + N N

H CH3 H CH3

At elevated temperatures, the broadening of the peaks followed by complete coalescence at

123°C was observed. 21 A 20.9 kcal/mol barrier for dimethylformamide isomer interconversion

is reported.20 In the case of N-methacryloyl-N-(phosphonomethyl)glycine, two factors are 26

governing the existence of two isomers: (1) hindered rotation around C-N bond which possesses

a partial double bond character and (2) additional stabilization owing to formation of two strong

intramolecular hydrogen bonds. One is formed between methacryloyl carbonyl oxygen and either

carboxylic or phosphonic hydrogen atoms and other spans between the carboxylic C=O and the

second OH group of the phosphonyl moiety as indicated on Figure 3.3. As revealed by the NMR

spectra (Figure 1.3), both conformers are equally populated and, therefore, have similar stability.

Figure 3.3. Two different isomers of N-methacryloyl-N-(phosphonomethyl)glycine.

It is interesting to note that chemical shifts of the methacryloyl double bond signals

appear at unusually high field (5.0-5.3 ppm). Typically, signals for this type of protons appear at

about 5.5-6.0 ppm downfield of TMS. This can be explained by the conformational preferences

of N-methacryloyl-N-(phosphonomethyl) glycine that are governed by steric factors (vide infra).

Similar to its methacrylated analog, N-acryloyl-N-(phosphonomethyl)glycine (AA) also

exists as a mixture of two slowly exchanging conformers at room temperature. Both 1H and 13C

NMR spectra of the acryloyl derivative (Figures 3.4 and 3.5) contain two sets of signals for all protons or carbons, respectively.

27

Figure 3.4. 1H NMR spectra of N-acryloyl-N-(phosphonomethyl)glycine (AA)

Figure 3.5. 13C NMR spectra of N-acryloyl-N-(phosphonomethyl)glycine (AA)

Apparently, the relative stability of two AA conformers is not the same as for MA and the ratio between them is about 1:2. The signals from the AA double bond hydrogen in the 1H 28

NMR spectrum appear at substantially lower field compared to those of N-methacryloyl-N-

(phosphonomethyl)glycine. In general, the electron donating effect of a methyl group is expected

to cause an upfield shift. For instance, the signals from the double bond terminal protons of

methyl acrylate appear at 6.40 and 5.82 ppm,22 while the parent signals for methyl methacrylate

are at 6.09 and 5.55 ppm. This represents a 0.3 ppm difference in chemical shifts. For

methacrylic and acrylic acids this difference is about 0.26 ppm. However, for methacrylic (MA)

and acrylic (AA) acid derivatives of glyphosate this difference is nearly 1.5 ppm. This large difference is unlikely to be entirely due to the electron donating abilities of the methyl group and can also be attributed to steric hindrance leading to conformational differences (vide infra).

Conformational Studies at Elevated Temperature

To further elucidate our experimental observations regarding hindered rotation around a partial

amide double bond in functionalized glyphosates, we have conducted the conformational NMR

studies at elevated temperatures. If our assumptions are correct, we expect to see a coalescence

of certain peaks in 1H NMR spectra of MA and AA. The peaks of interest for methacrylic acid derivative of glyphosate were those corresponding to two methyl groups at 1.90 and 1.95 ppm. It was found that increase in the temperature causes a broadening of those signals with complete overlap at 75°C (Figure 3.6). We observed the broadening for all other signals in the NMR spectra although coalescence occurred only for methyl and methylene (CH2-P) signals at

temperatures up to 85°C. Therefore, the barrier for conformer interconversion might be found by

implication of Eyring equasion.23

ΔG≠ = RT ln (κT / hk) (1)

where: R is a gas constant, T is a coalescence temperature, κ is a Boltzmann’s constant,

h is a Plank’s constant, and k is a rate constant. The rate constant can be found according to the 29

following expression:

k −= ννπ 21 2/)(2 (2),

where: ν1 and ν2 are the chemical shifts in Hz at the slow exchange limit.

Fitting the one equation into another yielded the free energy of activation (ΔG≠). Thus, for MA,

we found the activation energy for conformational isomerism at 348 K to be 17.3 kcal/mol.

85oC 75oC

70oC 65oC

O o OH P N CO H 60 C 2 OH O 3CH o 55 C H H MA 25oC

Figure 3.6. Temperature-dependent 1H NMR spectra of MA

In order to evaluate the barrier of internal rotation in the acrylated derivative of glyphosate, we have followed the changes in signals of methylene group attached to the phosphorous atom. At room temperature, the 1H NMR peaks of interest are represented by two 30

doublets (one doublet per conformer) centered at 3.95 and 3.98 ppm. Signals corresponding to

2 each of the conformers can be easily distinguished from the JHP = 12 Hz and the difference in

their intensity. Figure 3.7 follows the changes in the methylene signals obtained from the

temperature-dependent 1H NMR spectra of AA. At elevated temperatures, the signals of the two doublets broaden, come closer to one another, and at 65°C, they overlap to form one doublet

2 ≠ ( JHP = 12 Hz). Free energy of activation for C-N bond rotation (ΔG 338K) was determined

according to Eyring equation and similar to MA, was found to be 17.9 kcal/mol. Similar value

for the activation energy was found for MA conformational interconversion when we followed

≠ the methylene (CH2-P) signals (ΔG 358K = 17.8 kcal/mol).

65°C

H 60°C O H

OH P N CO2H OH O 55°C H H 45°C AA

35°C

25°C

Figure 3.7. Temperature-dependent 1H NMR spectra of AA 31

CHAPTER IV. POLYMERIZATION OF ACRYLATED AND METHACRYLATED

DERIVATIVES OF GLYPHOSATE.

In order to verify the ability of both derivatives of glyphosate to polymerize, thermal- and

photopolymerization experiments with UV- and visible light were performed. Both MA and AA

show a good solubility in water and a poor solubility in organic solvents. This presents additional

challenge, because most of common radical polymerization initiators are not water soluble. In

order to overcome this problem, we used water soluble diazo-initiators which are not very

efficient for photopolymerization because of low extinction coefficient at the wavelengths of

interest. An alternative was to introduce aliquots of common initiators dissolved in organic

solvents miscible with water (i. e. acetonitrile). Still, partial precipitation of initiator in the reaction mixture was observed for this approach. Thermal polymerization experiments were

conducted only for acrylated derivative of glyphosate. The following three initiators were used to

induce the polymerization (Table 4.1).

Table 4.1. Initiators for thermal polymerization.

Initiator Decomposition temperature/half life period VA – 057, Wako® 570C, 10 hours VA – 085, Wako® 850C, 10 hours t-Butyl peroxide 890C, 8 hours

After two hours of reaction, the product was isolated, dried and analyzed by IR and Mass

Spectrometry. The melting point of the homopolymer of acrylated glyphosate was found to be

around 115°C ± 10 with decomposition starting at ca. 200ºC. The infrared spectra of the

monomer and polymer were recorded and compared. The intensity of the amide signal increases

after the polymerization as represented on Figure 4.1. This effect can be explained by the loss of 32

conjugation of an acrylic double bond with an amide carbonyl group causing the intensity of

amide carbonyl group to increase.

I II

Figure 4.1. IR spectra of acrylated glyphosate (I) and poly(acrylated glyphosate) (II).

Photopolymerization experiments were carried out for both methacrylic and acrylic acid

derivatives of glyphosate. All experiments were carried out in deuterium oxide and double bond

conversion was monitored by 1H NMR. In order to induce the radical photopolymerization, we

tried several types of Norrish Type I initiators and water soluble diazo - initiators irradiated at

350 nm. Surprisingly, MA did not polymerize under these conditions despite the use of various initiators and long irradiation times.

On the other hand, the acrylic analog - AA - polymerized efficiently. Figure 4.2 shows

the conversion of AA double bond using 1.5 wt % Darocur 1173 initiator. Over 95% conversion

of double bond was achieved in less than 80 sec. The polymerization profile shows a typical

autoacceleration effect, leading to an increase of the initial polymerization rate, followed by an

autodeceleration effect. 33

Polymerization of Acrylamide 70mg AA 1.1mg Darocur 1173 (3 wt%) 100 0.8ml D2O @ 350nm NMR

80

60

40

20 Conversion [%] of Double Bond 0

0 20 40 60 80 100 120 Time [sec]

Figure 4.2. Photopolymerization of AA in the presence of Darocure 1173

-1 -1 (ε365 = 12 M cm ). UV spectra for Darocure 1173 adopted from ref. 24

The photopolymerization of AA initiated by Irgacure 651 was also investigated with the polymerization profile represented on Figure 4.3. Irgacure 651 has molar extinction coefficient at

350nm 6 times higher than that of Darocure 1173. Despite that, the irradiation of the reaction mixture for 7 minutes gave no more that 10% of double bond conversion.

Figure 4.3. Photopolymerization of AA in the presence of Irgacure 651. UV spectra for Irgacure

651 adopted from ref. 24. 34

The reason for such a low efficiency of polymerization is poor solubility of Irgacure 651 in water compared to that of Darocure 1173. As was mentioned before, to bypass the solubility problem we tried several diazo-water-soluble initiators. The use of diazo-initiators resulted in

efficient double bond conversion (90-95%) but required longer irradiation times (more than 15

minutes). These observations can be explained by the low value of molar extinction coefficient at

350 nm for these molecules. Values for molar extinction coefficients and efficiency of

polymerization for different initiators are summarized in table 4.2. Polymerization curves with

diazo-initiators are presented in Figure 4.4.

35

4H O NH CH3 CH3NH 2 PM of AA 1.5wt% WAKO Initiators N N CO H @ 350 nm (RVP) HO2C 2 100 N N CH3 CH3

80 VA 057 VA-085 VA-57 60 O CH CH O VA-086 3 3 OH N N OH 40 N CH CH N 3 3

20 VA 085 % Conversion of Double Bond of Double % Conversion

0 O CH3 CH3 O 0 2 4 6 8 101214161820 OH N N OH Time [min] N CH CH N 3 3

VA-086

Figure 4.4. Polymerization profiles for N-acryloyl-N-(phosphonomethyl)glycine with different initiators: VA 057, VA 085, VA 086.

Table 4.2 Summary of molar extinction coefficients and efficiency of polymerization for different initiators.

Initiator ε350nm Polymerization of AA M-1 cm-1

VA-057 16 slow (15min-95% conv.) VA-085 21 slow (15min-80-85% conv.)

VA-086 19 slow (15min-80-85% conv.)

Darocur 1173 40 100sec (95% conv.)

Irgacure 651 240 very slow (6min-10% conv.)

36

Polymerization induced by visible light was also implied for acrylic acid derivative of

glyphosate. For this purpose, the initiating system containing camphorquinone (CQ) and ethyl–4-

N,N-dimethylaminobenzoate (EDAB) synergist was utilized. As was reported previously25, CQ absorbs UV radiation in the region of 200-300 nm due to the π, π*transition and usually used as

efficient initiator for UV polymerization. But it also absorbs in the visible light region (400-500

nm) due to the n, π* transition of the α–dicarbonyl group and, therefore, frequently used as

initiator for visible light polymerization. 25-27 Addition of a tertiary amine to CQ enhances the

formation of free radicals capable of initiating the polymerization under exposure to visible

light.28 Figure.4.5 exhibits UV-spectra of camphorquinone. The measured extinction coefficient

-1 -1 28 of λ470 = 33 M cm corresponds closely to that from the literature.

CQ in hexanol-1 B 0.0012mol/l F 0.0017mol/l H 0.0033mol/l J 0.00512mol/l N 0.0086mol/l

0.5 Absorbance, o.u.

0.0

400 500 Wavelength, nm

Figure 4.5. Absorption spectra of camphorquinone in 1-hexanol

As in previous polymerization experiments, double bond conversion was followed by 1H

NMR. The dependence of double bond conversion as a function of time is represented in Figure

4.6. 37

70

60 50

40

30

20 10

Conversion of double % bond, double of Conversion 0 0 500 1000 1500 2000 Time, sec

Figure 4.6. Double bond conversion versus time in polymerization induced by visible light with

CQ/EDAB as an initiating system.

Visible light irradiation did not result in sufficient double bond conversion and displayed low rate of polymerization. The low extinction coefficient of CQ in 400-500 nm range may explain the low efficiency of polymerization.

The different photopolymerization behavior of methacrylic (MA) and acrylic (AA) acid derivatives can be explained by the different conformational preferences of the two compounds.

Methacrylic acid derivatives, although they polymerize more slowly than their acrylic acid counterparts, still often undergo efficient polymerization with a high degree of a double bond conversion.29 As discussed above, the signals from the double bond protons in the 1H NMR

spectrum of MA appear at unusually high field. Similarly, the double bond signals for N,N- 38

dimethylmethacrylamide appear at 5.3 and 5.03 ppm.22 We submit that steric bulk of the methyl group forces the MA molecule to adopt a conformation wherein the methacrylic C=O and C=C moieties are not coplanar. This prevents conjugation of the carbonyl group with the double bond, leading to deactivation of the double bond and its inability to undergo polymerization under free radical polymerization conditions. Other N,N-disubstituted methacrylamides have also failed to polymerize under similar conditions.29

The ability of the acrylated derivative of glyphosate to copolymerize was also

investigated. The 2-hydroxylethyl acrylate was selected as a comonomer because of its good

solubility in water and high reactivity in photopolymerization processes. The double bond

conversion for both monomers was monitored by 1HNMR. Results are represented in Figure 4.7.

According to polymerization profiles copolymerization occurs efficiently within two minutes

with 95% conversion of double bond for both monomers. Therefore, acrylated glyphosate proved

to polymerize and copolymerize efficiently with commercially available photoinitiators under

exposure by UV- light.

1.00

0.90

0.80

0.70 AA HEA 0.60

0.50

0.40 Extent of polymerization of Extent 0.30

0.20

0.10

0.00 020406080100120 Time (sec)

Figure 4.7 Copolymerization profile for acrylated glyphosate and 2-hydroxy ethylene acrylate 39

Molecular mass determination by SEC. The molecular mass of the poly(acrylated glyphosate) was estimated by size exclusion chromatography. Two asymmetric peaks corresponding to the high-molecular weight polymer and oligomer residues were observed. HPLC gel-filtration profile of homopolymer of acrylated glyphosate is presented on Figure 4.8.

Figure 4.8. SEC profile of poly(acrylated glyphosate)

The molecular mass of the fraction corresponding to the elution time of 20-35 min was determined to be about 55 kDa using the calibration line shown in Figure 4.9. The second peak on the chromatogram definitely belongs to the low molecular weight products.

40

Figure 4.9. Calibration curve for molecular mass determination of poly(acrylated glyphosate).

According to obtained data molecular weight of poly(acrylated glyphosate) belong to range of 55 kDa ± 10kDa. 41

CHAPTER V. COMPUTATIONAL STUDIES OF ACRYLATED AND METHACRYLATED

DERIVATIVES OF GLYPHOSATE.

The conformational preferences in both AA and MA were further probed by the calculations. Only one series was evaluated for rotation of the amide C-N bond for both AA and

MA, since whether the methacryloyl or acryloyl carbonyl can accept hydrogen bonds from the phosphoryl or the carboxyl group will depend on the value of the dihedral angle (Figure 1.5).

Two global minima were identified for dihedral angles of ca. 0º (P-OH…O=C HB) and ca. 180º

(COOH…O=C HB), with the dihedral angle defined as C(P)-N-C=O (a-b-c-d). High level geometry optimization for two isomers favors the 0o isomer by 2.2 and 3.7 kcal/mol for MA and

AA, respectively. The observed larger difference in stability of AA conformers is consistent with

the experimentally observed difference in the NMR signal ratios for MA (1:1) and AA (2:1).

Figure 5.1. Results of C(P)-N-C=C Amide bond rotational analysis for both AA and MA.

Our calculations show that the values of the barrier for rotation around C-N bond of MA

and AA are quite similar. This suggests that rotation around the C-N bond in MA to not be 42

hindered by the methyl group. For both MA and AA, the calculated value for the rotational

barrier is around 20 kcal/mole (Figure 5.1). This number is in a good agreement with the

experimental barrier of 17-18 kcal/mole determined from the elevated temperature NMR

experiments.

Two structural series were generated by rotating the vinyl group of the methacryloyl

moiety, with hydrogen-bond donation to the methacryloyl carbonyl from either the phosphoryl

group or the carboxyl group. Possible local minima were identified at ca. 135° and ca. 240° for

both series, with the dihedral angle defined as C=C-C=O (1-2-3-4) (Scheme 2). All identified

local minima had methacryloyl dihedral angles (1-2-3-4) of roughly ±120º (Figure 2.5). The two

lowest-energy minima found had calculated energies within 200 cal of each other and

methacryloyl dihedral angles of -123º and +121º.

Structural series’ were also generated by rotating the vinyl group of the acryloyl moiety

of AA, with hydrogen-bond donation to the acryloyl carbonyl from either the phosphoryl group

or the carboxyl group. Possible local minima were identified at ca. 0º, ca. 135° and ca. 240° for

both series, with the dihedral angle defined as C=C-C=O (1-2-3-4). Another possible local

minimum was identified at 60º for the series in which the carboxyl group served as the hydrogen-bond donor to the acryloyl carbonyl.

Identified AA local minima showed a wider range of both energy and dihedral angle. As with MA, the “global” minimum had the phosphoryl group acting as a hydrogen-bond donor to the acryloyl carbonyl oxygen. However, optimized dihedral angles fell into two groups: ±2º (the lowest-energy structures) and ±130º. Structures in which the acryloyl group was twisted ±130º lay about 3.5 kcal/mol above the “global” minimum (Figure 5.2). 43

Figure 5.2. Results of C=C-C=O acryloyl bond rotational analysis for both AA and MA.

These calculations confirm that the lowest energy conformations of the acryloyl and

methacryloyl groups (in AA and MA respectively) differ in the coplanarity of the vinyl and

carbonyl moieties as measured by the dihedral angle C(1)=C(2)-C(3)=O(4) [ϕ]. In AA, the vinyl and carbonyl groups are essentially coplanar (ϕ = 0°) while in MA they are non-coplanar (ϕ =

120°) (Figure 5.3). This non-planarity of the acrylic moiety in MA deactivates the double bond

which is consistent with the experimentally observed lack of polymerization for methacrylated

glyphosate. Figure 3.5 shows the optimized structures of the lowest energy conformations for

MA and AA.

Figure 5.3. B3LYP/6-31++G(d,p) optimized structures of the lowest energy isomers of

AA (left) and MA (right). 44

CHAPTER VI. BIOLOGICAL ACTIVITY OF ACRYLATED GLYPHOSATE AND

POLY(ACRYLATED GLYPHOSATE).

In order to investigate the biological activity of the novel compounds (acrylated

glyphosate and its homopolymer), we have conducted several types of biological experiments.

The model organism for the first series of experiments was E.Coli (the common organism of microbiological investigations).It is known that glyphosate affects the shikimic acid pathway in

algae, fungi, bacteria and higher plants, but there is still no uniform opinion about inhibition of

E.Coli growth by glyphosate.30,31 We were interested in obtaining the information regarding

toxicity of acrylated glyphosate toward this type of microorganisms. The first experiment was

done with E. Coli strain DH 5α in LB medium at 37C°. We had three samples of LB media with

culture; two of them were loaded with glyphosate and acrylated glyphosate with the final

concentrations of herbicides of 0.01M. Cell growth was monitored by measuring the optical

density of cultures at 600nm. No difference in growth of different samples of culture was

detected even after 6 hours. Two explanations can be suggested. First, because E. Coli is a

heterotrophic bacteria, it cannot be affected by glyphosate for physiological reasons. Second, LB

is a highly nutrient medium and its chemical composition cannot be determined. Therefore, it is

possible that some components of the medium simply destroy the additives. Also, the availability

of nutrients in excessive amounts makes organisms to propagate and mutate extremely

intensively and, possibly, an initial inhibition due to the presence of toxins was not detected. In

order to clarify this issue, we have changed the growth medium to M9 minimum medium which

possesses minimum amount of nutrients. This ensured us that additives would not be destroyed

during the course of experiment by the medium components and the rate of propagation will be

slow making it possible to detect a potential inhibition of bacteria growth by toxins. Strain, the 45

amount of herbicides in samples and method of growth detection remained the same. Growth

curves for the control and the samples loaded with glyphosate and acrylated glyphosate are

presented in Figure 6.1.

1.4 1.2 1 Control 0.8 Glyphosate 0.6 D, o.u. Monom er 0.4 0.2 0 0 5 10 15 20 25 Time, hrs

Figure 6.1. DH 5α growth curve

Significant inhibition of growth by glyphosate was observed during the first 7 hours. In

case of acrylated glyphosate, inhibition was also detected, but less than that for the glyphosate

and for the shorter period of time (5 hours). After 22 hours, the optical density for all samples was almost the same. This is an indication that suppression of growth had been somehow overcome by bacteria. After 25 hours, the optical density of the samples loaded with glyphosate and acrylated glyphosate was even bigger than that for the control sample. We repeated this experiment several times to ensure reproducibility of these results. These phenomena can be explained as follows: at the beginning, the bacterial growth is suppressed by the toxic shock caused by the presence of additives. After 5-7 hours, bacteria apparently produce some operons which enable them to resist. Finally, the intensive growth after 25 hours might be explained by 46

bacterial ability to use glyphosate and acrylated gloyphosate as sources of phosphorous (one of the most important nutrients). Most likely, this ability is due to production of σ-factors, which

allow the organisms to use toxins as nutrition source.

In order to confirm our observations, we have used the same strain of E. Coli and M9

agar plates loaded with glyphosate and acrylated glyphosate as a growing medium. Growth of

bacteria on pure M9 agar plates was used as a control. After 36 hours of propagation at 37°C, the

number of colonies on each plate was counted. The number of colonies on a control plate was

25. Count of 6 and 11 for plates loaded with glyphosate and acrylated glyphosate, respectively, was observed (Figure 6.2). Therefore, both toxins possess some antibacterial activity towards E.

Coli at the early stages. Although activity of the novel compound is less than that of glyphosate, it is still quite significant.

A (25) B (6) C (11)

Figure 6.2. Number of colonies on M9 agar plates loaded with glyphosate B, acrylated

glyphosate C, and pure M9 agar A.

Plates underwent further examination after 7 days (168 hours) of propagation at 37°C. An

evenly distributed layer of growing culture was found on all plates, with more intensive growth

observed for plates loaded with glyphosate and acrylated glyphosate. The results of the second 47 experiment confirmed earlier conclusions regarding the absence of the long-term toxicity toward

E. Coli.

The next step in our investigation was to test the toxicity and continuous action of acrylated glyphosate towards organisms of interest (photosynthesizing unicellular algae, cyanobacteria, diatoms) and compare to that of unfunctionalized glyphosate. As organisms of interest, cyanobacteria Synechocystos 6803 (fresh water strain), cyanobacteria Synechococcus

7002 (salt water strain), and green algae CD1Red were selected. The method of toxicity determination remained the same as for E. Coli growth with the exception that the optical density was monitored at 680 nm (chlorophyll a absorption). The test results for Synechococcus 7002 are represented on Figure 6.3 and the summary of results is listed in Table 6.1. For all three organisms, a significant inhibition of growth in the treated samples was detected. According to our observations, the toxicity of glyphosate and acrylated glyphosate towards photosynthetic microorganisms was similar. No growth within the treated samples was detected after 30 days of propagation. This confirms the irreversible nature of the inhibition of growth in photosynthetic organisms.

0.8

0.7

0.6

0.5 control 0.4 glyphosate AA 0.3 D, o.u. @ 680 nm

0.2

0.1

0 012345678 Days

Figure 6.3 Growth responce of Synechococcus 7002 to the treatment with glyphosate and acrylated glyphosate. 48

Table 6.1. Results of the Biological Activity Screening

Optical Density at 680 nm after 7 days of media Organism Description growth Control Glyphosate Acrylated Glyphosate (AA) CD1 Red Green Algae 0.55 0.02 0.02 Synechocystos 6803 Cyanobacteria 1.4 0.58 0.54 (fresh water strain) Synechococcus 7002 Cyanobacteria 0.72 0.04 0.07 (salt water (0.75 strain) NaCl)

Next, the critical question regarding the nature of the herbicidal activity of acrylated glyphosate had to be answered. Because the synthetic product (AA) contains ca. 30-40% of NaCl, the observed toxicity might be ascribed to the presence of a salt. A simple control experiment using the culture sample loaded with NaCl in the amount equal to that in the acrylated glyphosate proved this assumption to be wrong (Figure 6.4).

0.8

0.7

0.6

0.5 control salt 0.4 glyphoate AA 0.3 D, o.u. @ 680 nm

0.2

0.1

0 012345678 Days

Figure 6.4. Synechococcus 7002, salt sensitivity test

49

The experiment involving BG 11 agar plates showed that all tested microorganisms

(Synechocystos 6803, Synechococcus 7002 and CD 1 Red) do not grow on plates loaded with glyphosate or acrylated glyphosate. The complete inhibition of growth and bleaching of the area was observed compared to the control sample.

Homopolymer of acrylated glyphosate was synthesized and tested in order to elucidate its

herbicidal activity. The surface of a Petri dish was covered with homopolymer solution, solvent

was evaporated and this resulted in a thin film of AA homopolymer. This was covered with BG

11 agar layer. BG 11 agar plate without additives was used as a control. CD 1 red was inoculated

on both plates. After 3 days, the complete inhibition of an algal growth was observed on a plate loaded with the homopolymer of acrylated glyphosate (Figure 6.5). Similar results were obtained

for all other cultures examined. These results suggest that the homopolymer of AA also

possesses herbicidal activity toward cyanobacteria and algae.

Figure 6.5. Herbicidal activity test for the agar medium placed over the layer of AA

homopolymer (right dish) and control (left dish).

In order to further investigate the herbicidal activity we carried out Kirby-Bauer

susceptibility tests. These tests are commonly utilized in molecular microbiology. These 50

experiments aided us in achieving the following two tasks: (1) elucidation of the minimum

concentration of novel compound to inhibit growth of culture and comparison of that with the

concentration of a glyphosate and (2) finding out what kind of organisms is more resistant to the

treatment by herbicides.

In addition to the target organisms used in previous experiments, we have tested 2 strains

of diatoms – Phaeodactylum tricornutum and Navicula pelliculosa. These are photosynthetic

organisms known to initiate one of the first steps in biofouling.32-35 Another culture adapted for

these experiments was red algae LS 0504. Typical images of Kirby-Bauer test results are

depicted in Figure 6.6. summary of the results is presented in Table 6.2.

A). . B).

Figure 6.6 Typical images of Kirby-Bauer test results. A. inhibition zones produced by acrylated glyphosate (0.3M), B. inhibition zones produced by homopolymer of acrylated glyphosate

(0.4g/ml).

Based on the results of Kirby-Bauer tests, it is clear that the acrylated glyphosate

possesses reasonable herbicidal activity which is somewhat less than that of a glyphosate but

much higher than that of control acrylamide. Homopolymer of acrylated glyphosate also

demonstrates the reasonable herbicidal activity. Minimum concentration of acrylated glyphosate required for inhibition varies between 0.1-0.15 M, depending on the tested organism. According to obtained results, green algae CD1 Red are the most resistant organism. The most important 51 result is that the homopolymer of acrylated glyphosate is active against diatoms which are target organisms in marine biofouling sequence.

Table 6.2. Average radii of the inhibition zones measured around the filter paper circles treated with the designated compound. Culture Herbicide Inhibition by Inhibition by polymer, mm Inhibition by acryl. concent. mol/L glyphosate, mm glyph., mm

Syn 7002 0.0003 No 9 mm (0.025g/ml) No

- 0.003 6 13.5 mm(0.25g/ml) No

- 0.03 8 5.5

- 0.3 Bleaching 15

CD1 Red 0.0003 no 13 mm (0.025g/ml) No

0.003 no 22 mm (0.25g/ml) No

0.03 7.5 No

0.3 Bleaching 10

LS 0504 0.0003 No 10 mm (0.025g/ml) No

0.003 No Bleaching 0.25g/ml) No

0.03 8 No

0.3 Bleaching 16

Phaeodactilum d-om 0.003 No No (0.0045g/ml) No

0.03 8 No (0.045g/ml) No

0.3 10 11.5 mm (0.45g/ml) 11

Navicula d-om 0.003 No No (0.0045g/ml) No

0.03 No No (0.045g/ml) No

0.3 14 0. 12 mm (0.45g/ml) 13

52

CHAPTER VII. COATING FORMULATION, CHARACTERIZATION AND

ELUCIDATION OF HERBICIDAL ACTIVITY.

In order to elucidate the herbicidal activity of acrylated glyphosate incorporated into the

model acrylic coating formulation (459S2), we have synthesized pellets loaded with 0.3M,

0.15M and 0.1M of acrylated glyphosate and glyphosate, respectively. There is a principal

difference between these two types of formulations. For the pellets loaded with acrylated

glyphosate, we expect copolymerization and incorporation of a novel compound into the

polymer backbone. For glyphosate copolymerization is impossible because of the absence of a

proper functionality, glyphosate can only be distributed in the pores of acrylic matrix as an inert

biologically active filler. In order to confirm our assumption, we introduced a coating release

experiment. It is worth notice that acrylated glyphosate as well as glyphosate sufficiently soluble

in water. Because both glyphosate and acrylated glyphosate display good solubility in water, we

kept both pellets in 1 ml of D2O for the period of 3 weeks. Periodically, the pellets washings

were analyzed by 1H NMR. Results of the coating release experiments are presented in Figure

7.1.

100

90

80

70

60 % 459S2+glyph 50 459S2+Acr.G

Release, 40

30

20

10

0 0 5 10 15 20 Days

Figure 7.1. Normalized amount of active component released from the coatings as a function of testing time. 53

Almost complete release of unfunctionalized glyphosate (94%) was observed after 18

days of testing while acrylated glyphosate remained in the pellets. These results confirm that

acrylated glyphosate copolymerized with components of acrylic formulation.

Further, we have made several attempts to analyze the composition of the pellets loaded

with acrylated glyphosate and to find additional evidence for copolymerization of the acrylated

glyphosate. Several difficulties arose during the course of these investigations. First, the acrylic

polymer is highly cross linked and, therefore, insoluble in organic solvents. The second problem

was due to low content (close to the detection limit of most of devices) of acrylated glyphosate in

the pellet. First, we tried to use FTIR spectroscopy to observe any differences between the pure

resin, resin loaded with AA, pure AA and homopolymer of AA. The results obtained were

insufficient. We observed no changes between spectra of the pure and the loaded resins. Neither

did no detect any characteristic peaks of acrylated glyphosate in the spectra of the loaded resin.

Apparently, the reason of our lack of success is the low concentration of an additive in the analyzed probe.

The second attempt to evaluate the composition of resins was based on the solid state

NMR technique. Spectra of pure resin, resin loaded with acrylated glyphosate and glyphosate

were obtained and compared. Results of these experiments are presented in Figure 7.2

54

Pure resin Resin+AA

Resin+Glyphosate

Figure 7.2 Solid state NMR spectra for pure acrylic formulation and formulations loaded with

acrylated glyphosate or glyphosate respectively.

Comparison of spectra shows broadening and coalescence of peaks for resins loaded with

AA and glyphosate in the region between 110 and 140 ppm. This broadening can be explained only by presence of glyphosate and its derivative. However, the presence or absence of the double bond of an acrylated glyphosate in the corresponding resin is not obvious because of the overlap with peaks belonging to the resin’s aromatic components. 55

An additional attempt to evaluate the copolymerization of an acrylated glyphosate was made by

using MALDI-ToF-MS. Results of experiment are shown in Figure 7.3.

Figure 7.3 MALDI-ToF-MS chromatogram for model resin loaded with AA (0.3M).

Unfortunately, MALDI mass spectrum yielded no useful information because of the poor

solubility of highly cross-linked resin in any solvent. As reported, the inhomogenity between the matrix and the sample leads to direct ionization of the polymer. As a result, the polymer fragments are not fully ionized leaving ion levels below the detection limit.36 What we have seen

in the mass spectrum, are only the peaks due to presence of low molecular weight oligomers,

which are soluble in organic solvent (THF). The increment of 44 m/z between peaks is likely due to polyethylene glycol residues. 56

The last attempt to identify the presence of acrylated glyphosate in the backbone of the polymer was based on a Scanning Electron Microscopy. We are interested in the resin surface changes before and after treatment with water. In order to do that, we have coated glass slides with pure resins and resins loaded with glyphosate and acrylated glyphosate, respectively.

Pictures of both freshly prepared coating surfaces and surfaces after 13 days of water immersion were taken by SEM. Results of the experiments are given in Figure 7.4. 57

0 days in water after 13 days in water

Pure resin

Resin loaded with glyphosate

Resin loaded with acrylated glyphosate

Figure 7.4 Images of resin surfaces before and after treatment with water for 13

days 58

Apparently, the surface disruption of the glyphosate containing resin is due to dissolution

of glyphosate and its release from the resin. These results might serve as a preliminary prove that

acrylated glyphosate indeed copolymerizes with the components of the model acrylic

formulation. If this was not true, we would have to observe the surface disruption of the resins

loaded with acrylated gyphosate, because of almost equally good solubility of both herbicides in

water. Also, as was previously showed, we observed efficient copolymerization of acrylated

glyphosate with 2-hydroxy ethyl acrylate. This fact might be considered as an indirect evidence

of acrylated glyphosate incorporation into the backbone of the model acrylic resin.

The next step was to test the biological activity of AA incorporated into a model acrylic coating formulation 469S2. As controls, the pure acrylic resin and resin loaded with glyphosate

were used. The herbicidal activities of all three formulations were compared. Two different

strains of diatoms were used as target organisms for coating testing. The results are presented in

Table 7.1. As can be seen from the table, similar observations could be made for both diatom

strains tested. When acrylic resin contains no herbicide, only slight or no inhibition of diatom

growth is observed. Incorporation of similar molar concentrations of non-functionalized

glyphosate and acrylated glyphosate AA into 459S2 resin results in either substantial growth

inhibition or complete bleaching in the case of high concentrations of herbicide. This is observed

despite a different arrangement of herbicide within the coating. The former is a poly acrylic

matrix with non-functionalized glyphosate particles trapped in the pores while the latter is a

copolymer with acrylated glyphosate functionality chemically incorporated into the polymer

backbone. 59

Table 7.1. Average radii of the inhibition zones measured around the polymer pellets containing

the designated compound.

Strain name Herbicide in Concentration in Diameter of inhibition zone, mm (diatom) polymer tablet tablet, M (% mass) Navicula None 0.0 (0.0) 10 with slight bleaching AA 0.15 (3.2) 20 AA 0.3 (6.0) bleaching Glyphosate 0.15 (2.4) 22.5 Glyphosate 0.3 (5.0) bleaching Phaeodactilum None 0.0 (0.0) No inhibition AA 0.1 (2.4) 23 AA 0.3 (6.0) bleaching Glyphosate 0.1 (1.6) 17.5 Glyphosate 0.3 (5.0) bleaching

This leads to an important observation: in the case of AA copolymer, the glyphosate functionality is chemically incorporated into a polymer which prevents its release but the resulting polymer still retains significant biological activity. This observation highlights an

excellent potential of acrylated glyphosate (AA) as biologically active component of the

antifouling coatings. 60

Conclusions

One of the most efficient ways to prevent biofouling is to incorporate the biocide into the

polymer coating. The potential biocide should meet several requirements. It should be (a)

inexpensive and commercially available, (b) minimally toxic towards marine animals and

humans and (c) biodegradable. Because glyphosate is commercially available, cheap and

biodegradable (by many strains of soil bacteria) and the mechanism of glyphosate action is well

studied, this compound might be used as a potential biocide for biofouling. In the present work

the synthesis and characterization of potential glyphosate-based biocide – acrylated glyphosate -

was reported for the first time.

To incorporate glyphosate into the coating we need to introduce the proper functionality

into the molecule. According to the synthesis reported in the current work, it is possible to

produce acrylated and methacrylated glyphosates with up to 95% yield for batches up to 40 g

magnitude of acrylic analog.

1H NMR characterization indicated that the novel compound exists as a mixture of two

isomers because of the partial character of the carbonyl double bound. Application of high

temperature 1H NMR allowed us to confirm this idea and found the energy barrier for free

≠ rotation to be ΔG 338K = 17.9 kcal/mol. Computation studies allowed us to elucidate the optimized conformations for both isomers

Thermal and photopolymerization as well as copolymerization experiments showed that

the novel acrylated glyphosate is able to polymerize and copolymerize efficiently under standard conditions. 61

Biological activity tests demonstrated that both acrylated glyphosate and poly(acrylated glyphosate) (!) possess herbicidal activity which although slightly less than that of the parent glyphosate but still significant if referred to controls.

Incorporation of acrylated glyphosate into a model acrylic formulation yielded a polymer

with acrylated glyphosate chemically blended into the backbone of acrylic photopolymer.

Coating release experiments showed no leaching of acrylated glyphosate out of the acrylic matrix confirming the completion of copolymerization. To our surprise and excitement, biological tests demonstrated high inhibitory activity of acrylic formulations loaded with acrylated glyphosate toward microorganisms of interest, whereas pure acrylic formulations with no additive showed slight or no inhibitory activity at all. Thus, acrylated glyphosate remains toxic against target organisms even when chemically incorporated into the polymer matrix.

Therefore, we conclude that the novel acrylated glyphosate is a promising biocide for new antifouling coatings. 62

References.

1. Callow, M.E., Callow J.A., Biologist 2002, 49, 1

2. Rittschof, D., Biofouling 2000, 15, 119

3. Marine BioFoulNet http://www.biosciences.bham.ac.uk/external/biofoulnet

4. Callow, M. E., Chem. Industry 1990, 5, 123.

5. Yebra D.M.; Kiil, S.; Dam-Johansen, K. Prog. Org. Coat. 2004, 50, 75

6. Woods Hole Oceanographic Institution (WHOI), US Naval Institute, Annapolis, Iselin,

COD 1952

7. Gitlitz M.H. J. Coat. Tech. 1981, 53, 46

8. Brandy R.F. Prog.Org. Coat. 2001, 43, 188

9. Bentley, R. Biochemistry and Molecular Biology 1990, 25, 307

10. Levin, J. G.; Sprinson, D. B. Journal of Biological Chemistry 1964, 239, 1142

11. Schönbrunn, E. , Eschenburg, S. , Shuttleworth, W. A. , Schloss, J. V. , Amrhein, N. ,

Evans, J. N. S. & Kabsch, W. Proc. Natl. Acad. Sci 2001, 98, 1376

12. http://umbbd.ahc.umn.edu/gly/glymap University of Minnesota

13. Gough, S. T. D. U.S. Patent 4,364,767, 1982.

14. Gaussian 03, Revision C.02, Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G.

E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, Jr., J. A.; Vreven, T.; Kudin, K. N.;

Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi,

M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.;

Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.;

Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.;

Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; 63

Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;

Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain,

M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.;

Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.;

Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,

M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.;

Chen, W.; Wong, M. W.; Gonzalez, C.; and Pople, J. A.; Gaussian, Inc., Wallingford CT,

2004.

15. Ma, J.; Xu, L.; Wang, S. Weed Science 2002, 50, 555.

16. Bauer A.W.; Kirby W.M.; Sherris J.C. Am. J. Clin. Pathol. 1966, 45, 493

17. Knuuttila, P.; Knuuttila, H. Acta Chem. Scand. 1979, 33, 623

18. Riebel, H.-J. PCT Int. Appl. 9731534, 1997

19. Miyata, H. K.; Sasaki, T. F.; Kohei Morikawa, K. U.S. Patent 5,859,289, 1999

20. Eliel, E. L.; Wilen, S. H.; Mander C. N. Stereochemistry of Organic Compounds, John

Wiley & Sons, Inc.: New York, 1994.

21. Silverstein, R.; Webster F. Spectrometric Identification of Organic Compounds, Sixth

Edition, John Wiley & Sons, Inc.: New York, 1997.

22. http://www.aist.go.jp/RIODB/SDBS/cgi-bin/cre_index.cgi, Spectral database System.

23. Ionin, B.; Ershov, B.; NMR Spectroscopy in Organic Chemistry, Plenum Press: New

York, 1970.

24. www.ciba.com Database of photoinitiators

25. Evans, T.R., Leermakers, P.A., J Am Chem Soc 1967, 89, 4380

26. Charney, E., Tsai, L., J Am Chem Soc 1971, 93, 7123 64

27. Luk, C.K., Richardson, F.S., J Am Chem Soc 1974, 96, 2006

28. Yu, Q., Nauman, S., Santerre, J.P., Zhu S., J Appl Polym Sci 2001, 82, 1107

29. Sokolova, T.; Chetyrkina, G. Vysokomolekulyarnye Soedineniya 1961, 3, 244.

30. Fisher, R.S.; Berry, A.; Gaines, G.; Jensen, R.A., J.Bacteriol. 1986, 168, 1147

31. Park, H.; Helsenbeck, J.L.; Kim, H.J.; Shuttleworth, W.A.; Pack, Y.H.; Evans, J.N.;

Kang, C.H., Molecul. Biol. 2004, 51, 963

32. Sitaraman, K.; Wang, N.; Ober, C. K.; Finlay, J. A.; Callow, M. E.; Callow, J. A.;

Hexemer, A.; Sohn, K. E.; Kramer, E. J.; Fischer, D. A. Biomacromolecules 2006, 7,

1449.

33. Finlay, J. A.; Callow, M.; Ista, L.; Lopez, G. P.; Callow J. A. Integr. Comp. Biol. 2002,

42, 1116

34. Geesy, G. G.; Wigglesworth-Cooksey, B.; Cooksey, K. E. Biofouling 2000, 15, 195

35. Callow, M. E. In Biofilm: Recent Advances in Their Study and Control; Evans, L. V.,

Editor; Harwood Academic Publishers: Amsterdam, 2000; p. 189.

36. Bahr, U.; Deppe, A.; Karas, M.; Hillencamp, F. J. Anal. Chem., 1992, 64, 2866