A Remotely-Controlled, Semi-Automatic Target System for Rutherford Backscattering 1 Spectrometry and Elastic Recoil Detection Analyses of Polymeric Membrane Samples 2 3 P.J. Attayeka,b, E.S. Meyera, L. Lina, G.C. Richc,d, T.B. Cleggc,d, O. Coronella,* 4 aDepartment of Environmental Sciences and Engineering, Gillings School of Global Public 5 Health, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7431 6 bDepartment of Biomedical Engineering, University of North Carolina at Chapel Hill, Chapel 7 Hill, NC 27599-7575 8 cTriangle Universities Nuclear Laboratory (TUNL), Durham, NC 27708-0308 9 dDepartment of Physics and Astronomy, University of North Carolina at Chapel Hill, Chapel 10 Hill, NC 27599-3255 11 *

13

Corresponding author [tel:+1-919-966-9010; fax:+1-919-966-7911; e-mail: coronell@unc.edu] 12

Abstract: A new target system for Rutherford backscattering spectrometry and elastic recoil 14

detection analysis is described which enables remotely controlled, semi-automatic analysis of 15

multiple organic polymer samples without exceeding damaging incident beam fluences. Control 16

of fluence at a given beam current is achieved using two stepper motors to move a thin 17

aluminum disc loaded with polymer samples both radially and azimuthally across the beam. 18

Flexible beam spot locations and sample irradiation times are remotely controlled in two steps 19

via two custom LabVIEWTM programs. In the first step, a digital image of the target disc is 20

converted into precise radial and azimuthal coordinates for each mounted polymer sample. In 21

the second step, the motors implement the user-directed sample irradiation and fluence. 22

Schematics of the target system hardware, a block diagram of interactions between the target 23

system components, a description of routine procedures, and illustrative data taken with a 2-MeV 24

4He2+

Key Words: Rutherford backscattering spectrometry, elastic recoil detection, membranes, thin 26

film composite, reverse osmosis, nanofiltration 27

analysis beam are provided. 25

28

29

1. Introduction 30

31

Rutherford backscattering spectrometry (RBS) has recently been used as an analytical technique 32

to study the physico-chemical properties of the active layers of thin-film composite membranes 33

for water purification [1-9]. Thin-film composite membranes used in applications such as 34

reverse osmosis and nanofiltration consist of an ultrathin (~50−200 nm) organic active layer that 35

lies on top of an intermediate polysulfone support (~30 µm) backed by a polyester fabric (~200 36

µm) [10, 11]. The performance of such thin-film composite membranes, in terms of water 37

permeation and contaminant rejection, is mostly determined by their active layer [10, 11]. As a 38

result, the development of new improved membranes significantly benefits from studying the 39

physico-chemical properties of the active layers and their interactions with water and 40

contaminants. The recently published [1-6, 8] membrane characterization procedures that use 41

RBS for sample analysis take advantage of the ability of RBS to resolve the active layer from 42

their significantly thicker polysulfone and polyester supports [2, 5, 12]. 43

44

During RBS analyses, a high-energy ion beam (e.g., 2-MeV 4He+ [2, 5]) is used to irradiate the 45

target sample, and the spectrum resulting from the backscattered ions is analyzed to obtain the 46

elemental composition of the irradiated sample with nanometer depth resolution [2, 5, 12]. 47

Unfortunately, beam irradiation of organic polymers results in polymer degradation [13-17] with 48

a corresponding emission of beam-generated small molecules [13, 14] that leads to a changing 49

elemental composition of the irradiated sample as a function of beam ion fluence [13-15, 17]. 50

Studies [13, 16, 17] of polymer degradation upon beam irradiation have shown, however, that 51

induced changes in the elemental composition of irradiated samples can be neglected below 52

fluence thresholds that depend on the polymer analyzed. Among the materials relevant to the 53

structure of thin-film composite membranes, polysulfone has shown the lowest fluence threshold 54

value at 3×1014 ions/cm2 using a 2-MeV 4He+

56

ion beam [17]. 55

For RBS to be useful in the study of the properties of the active layers of thin-film composite 57

membranes [1-9], the beam ion fluence is maintained below the threshold values where changes 58

in the elemental composition of the sample analyzed would be detected. Given that the elements 59

(i.e., H, C, O, N, S) that make up thin-film composite membranes have relatively small scattering 60

cross sections (0-342 mb/sr) [12], the only possible strategy to avoid exceeding fluence threshold 61

values, while obtaining enough counts to quantify accurately the sample elemental composition, 62

is to scan a relatively large area (e.g., a few square centimeters) with the analysis beam. One 63

achieves this either by analyzing multiple points of the membrane sample or by continuously 64

scanning the ion beam over the sample [2-8]. Such scanning patterns have been achieved by 65

continual manual positioning of scattering target stages, a technique which is highly impractical 66

for analysis of multiple samples. Active layers of thin-film composite membranes can also be 67

studied after isolating the active layer on a solid surface by peeling off the polyester backing and 68

dissolving the polysulfone support using organic solvents [18-21]. Isolated active layers are 69

generally relatively small (i.e., a few square centimeters) and irregular in shape, and therefore, 70

the RBS analysis of isolated active layers requires both irregular scanning patterns and 71

millimeter-accurate positioning of the beam over the sample to stay within the sample 72

boundaries. 73

74

Thus, an innovative target system is needed to implement practical analysis of multiple organic 75

membrane samples while satisfying the fluence threshold, sample location, and scanning pattern 76

requirements described above. Accordingly, the objective of this work was to design, fabricate 77

and test a scattering target system for RBS analysis, and the sister technique of elastic recoil 78

detection (ERD) analysis, with the following capabilities: (i) remote control; (ii) irradiation of 79

samples according to regular and irregular scanning patterns defined by the user; (iii) positioning 80

of beam on target with millimeter-scale accuracy; and (iv) accommodation of multiple samples 81

of any shape on the sample holder. 82

83

2. Accelerator and Scattering Chamber 84

85

The new RBS/ERD target system utilizes systems associated with the tandem electrostatic 86

accelerator at the Triangle Universities Nuclear Laboratory (TUNL). Existing hardware at 87

TUNL includes a duoplasmatron ion source equipped with a sodium charge-exchange canal used 88

to produce a 30-keV 4He– ion beam which is injected into the accelerator. After initial 89

acceleration to the electrostatic accelerator terminal at 0.67 keV, the 4He– beam is stripped of its 90

electrons in a thin carbon foil inside the terminal, and then further accelerated. The emerging 2-91

MeV 4He2+ beam is momentum selected by deflection (52 degrees) through an analyzing magnet, 92

focused, and transported 15 meters to a multipurpose, aluminum scattering chamber. The 93

chamber is cylindrical in shape with internal diameter and depth of 59.7 cm and 26.0 cm, 94

respectively. The beam enters the chamber through entrance slits of adjustable aperture that 95

typically define a square 3-mm × 3-mm beam cross section. The beam spot on target is located 96

6.4 cm above a bottom plate that holds several detectors of scattered or recoil particles emerging 97

from the polymer film being used as a target. This bottom plate can be rotated while under 98

vacuum around the vertical axis of the chamber to set the desired detector angles. Signals from 99

the detectors are first sent through pre-amplifiers at the scattering chamber before being sent to 100

the accelerator control room for further processing and digitization. 101

102

3. RBS/ERD Target System 103

104

The new RBS/ERD target system consisting of a target rod, static stage, dynamic stage, and 105

target wheel is shown in Figure 1. The target rod secures and locates the rest of the target system 106

precisely inside the scattering chamber and allows for rotation of the system around the central 107

vertical axis of the chamber. The static stage houses a stepper motor that controls vertical 108

movement of the dynamic stage along two precisely located steel rods. The target wheel is 109

attached to a second stepper motor on the dynamic stage which controls its azimuthal motion. 110

The two stepper motors thus move the target wheel vertically and rotate it around its central axis. 111

The target system is controlled via two custom LabVIEWTM

119

programs. One program, 112

WheelScan, combines a digital picture of a loaded target wheel with input from the user to 113

establish the movement patterns of the two stepper motors. A second program, MotorLord, then 114

uses data generated by WheelScan to control the two stepper motors, thus implementing desired 115

movements of the target wheel. Figure 2 illustrates the interaction between the hardware, 116

software and user of the target system. Detailed information about all system components is 117

included in the following sections. 118

Insert Figure 1. Size suggestion: 1.5 column 120

Insert Figure 2. Size suggestion: 1.5 column 121

122

3.1. Hardware 123

124

3.1.1. Sample Holder Target Rod 125

126

The nickel-plated brass target rod supporting the sample holder system slides vertically and 127

rotates easily while maintaining a secure vacuum seal with the scattering chamber lid. The 128

positioning handle of the target rod has azimuthal fiducial marks in 5 degree increments to 129

indicate the angle of incidence of the beam with respect to the normal of the target wheel. The 130

positioning handle is also pinned to a frame atop the scattering chamber to support the weight of 131

the sample holder system and align the vertical axis of the rod with the horizontal beam axis. 132

Using an optical transit, the horizontal rotation axis of the target wheel was confirmed to 133

intersect both the axis of the beam and the vertical axis of the rod through the center of the 134

scattering chamber to within 0.5 mm. 135

136

3.1.2. Target Wheel 137

138

The aluminum target wheel, having 15.9-cm outer diameter and 0.8-mm thickness, is designed to 139

be lightweight and to hold as many membrane samples as possible while fitting within the 140

scattering chamber. The 1.59-cm inner diameter of the target wheel fits snugly over a Delrin® 141

hub to define its axial location while providing electrical insulation from the rest of the sample 142

holder apparatus. Four calibration holes, 0.5 mm in diameter, positioned in a 110-mm × 110-mm 143

square grid centered on the axis of the wheel are used as reference points to define the exact 144

position of membrane samples as described in Section 3.2.1. Additionally, the target wheel has 145

an engraved 3.18-cm diameter circle centered on the wheel axis and an engraved chord whose 146

closest point is located 7.6 cm from the center of the target wheel. Samples must lie within these 147

engraved boundaries as the vertical scanning range of the wheel is limited by the size of the 148

scattering chamber. The Delrin®

154

hub at the center of the wheel can easily be detached from and 149

re-attached to the sample holder, allowing for sample loading and removal. Membrane samples 150

may be secured to the wheel by double-sided conductive tape, which inhibits charge build-up 151

and conducts heat to the wheel when beam strikes the membranes, or alternatively by using two 152

perpendicular diametrical arrays of small threaded holes in the wheel to clamp samples in place. 153

3.1.3. Lower Stages 155

156

Two stages, one static and one dynamic, which are attached to the bottom of the target rod house, 157

separate identical vacuum-rated stepper motors (NEMA 11 bipolar vacuum-rated stepper motor, 158

Lin Engineering, Morgan Hill, CA, USA). The motor on the static stage controls the vertical 159

motion of the target wheel, while that mounted on the dynamic stage controls its azimuthal 160

rotation. 161

162

The shaft of the vertical motor is coupled directly to an Acme threaded rod (0.375-inch diameter, 163

16 pitch) and the corresponding threaded insert is attached to the dynamic stage. Motion of the 164

dynamic stage is guided by two steel rods, 6.3 mm in diameter, which fit into linear Teflon® 165

bushings. The stepper motor normally has 200 full steps per revolution (1.8 deg/step); however, 166

a motor driver/controller circuit (TMCM-310/SG, Trinamic Motion Control GmbH & Co. KG, 167

Hamburg, Germany) is used to microstep the motor eight times per full step resulting in more 168

precise control of the motion with 1,600 microsteps per revolution (0.225 deg/microstep). A 169

microswitch attached to the static stage indicates when the dynamic stage reaches its uppermost 170

point, giving the motor driver/controller a reference point for positioning the vertical motor. As 171

long as power to the motor driver/controller circuit is maintained, the driver/controller tracks the 172

number of microsteps from the reference point allowing for accurate vertical positioning of the 173

wheel. 174

175

The azimuthal motor drive shaft attaches to the Delrin® hub of the target wheel via a custom 176

keyed, shouldered shaft extension. The Delrin®

187

hub is pressed onto the shaft extension and 177

snuggly against the shoulder for precise positioning of the target wheel within the scattering 178

chamber. The rotary stepper motor has 200 full steps per revolution but is micro-stepped 64 179

times per full step resulting in 12,800 microsteps per revolution (0.0281 deg/step) allowing for 180

very precise azimuthal rotation of the target wheel. The azimuthal motor also has a small rear 181

drive shaft where an absolute magnetic encoder (MAE3 Absolute Magnetic Kit Encoder, US 182

Digital, Vancouver, WA) is attached to enable highly accurate tracking of the azimuthal position 183

of the target wheel. The encoder outputs a digital clock signal of 250 Hz whose variable duty 184

cycle is correlated to the azimuthal position of the drive shaft. The pulse width of the signal has 185

12-bit precision giving the encoder a resolution of 4,096 distinct duty cycles per revolution. 186

A graphite brush housed in Delrin® is attached to the dynamic stage of the target system. This 188

brush is spring loaded to make good sliding contact against the insulated target wheel during 189

rotation. A wire connected to the brush carries current outside of the scattering chamber for 190

remote beam current integration and monitoring. 191

192

Electronics outside the scattering chamber power the stepper motors, the absolute position 193

encoder, and the motor driver/controller. Electronic connections to the two stepper motors, 194

absolute encoder, and microswitch from the external power supplies and motor driver/controller 195

are established via a 25 pin D-subminiature (DB-25) vacuum feedthrough in the scattering 196

chamber wall. 197

198

3.2. Software 199

200

National Instruments LabVIEWTM and Vision Development Module programs are used to 201

control the target system. Two LabVIEWTM

205

virtual instruments called WheelScan and 202

MotorLord are used to control the beam irradiation of samples on the target wheel as described 203

in the following sections. 204

3.2.1. WheelScan 206

207

The first virtual instrument, WheelScan, determines the coordinate positions of the membrane 208

samples on the target wheel and the method and pattern for beam irradiation of each sample. 209

The primary input to WheelScan is a digital photograph of the target wheel loaded with 210

membrane samples, taken with the engraved target wheel chord oriented approximately 211

horizontally and located in the upper third of the picture. When WheelScan is executed, it 212

prompts the user to specify the width of the expected square beam spot in millimeters as well as 213

the expected incident angle. The user then selects the digital photograph of the loaded target 214

wheel and identifies the location of the four calibration holes on the wheel image by clicking on 215

them. Since the physical locations of the calibration holes on the wheel are precisely known, 216

WheelScan uses their pixel locations on the digital wheel image to determine the physical 217

dimension of an image pixel. WheelScan then converts locations in the image, specified by pixel 218

coordinates, into physical locations on the wheel itself, specified by polar coordinates (mm, 219

radians) with the origin at the center of the target wheel. 220

221

The user then specifies the number of individual samples to be analyzed, and designates the 222

scanning mode to be used for each sample by selecting between point, discrete rastering and 223

continuous rastering. For point samples, the user selects a single coordinate where the beam will 224

be focused for the duration of the exposure of the sample. Point samples are used for calibration 225

of RBS/ERD energy spectra, and should be samples that are not damaged by prolonged beam 226

exposure. For discretely- or continuously-rastered samples, the user employs a polygon tool to 227

draw on the digital wheel image a closed outline of the area to be analyzed. Using the specified 228

beam width, WheelScan decomposes the specified polygon area into a series of adjacent arcs that 229

define the beam irradiation pattern. For each arc, both the start and end angles and the radial 230

displacement are saved by the software. If the sample is to be discretely rastered, WheelScan 231

decomposes each arc into a series of discrete points that are spaced far enough apart to avoid 232

overlapping beam spots. After defining all samples, the user assigns a name to the file in which 233

WheelScan stores the data that define the raster patterns of all samples in the wheel. The 234

generated file is then ready to be used by the second virtual instrument, MotorLord. 235

236

3.2.2. MotorLord 237

238

The second virtual instrument, MotorLord, uses the data created by WheelScan to control the 239

two stepper motors on the target system and data acquisition from the RBS/ERD detector(s). 240

LabVIEWTM interfaces with the microswitch in the target stage and the electronics that comprise 241

the RBS/ERD data acquisition system (DAQ) via a small USB DAQ (USB NI-6009, National 242

Intruments, Austin, TX, USA) that handles digital, transistor-transistor logic input and output. 243

LabVIEWTM communicates with the motor driver/controller using a TCP connection established 244

via an Ethernet-to-RS232 converter (ESL 1 port RS-232 DB9, Lava Computer MFG Inc., 245

Toronto, Canada). The output signal from the absolute magnetic encoder is input into a 246

microcontroller (PIC18F4520, Microchip Technology Inc., Chandler, AZ) with a 20 MHz crystal 247

which calculates the duty cycle of the system. The microcontroller communicates with 248

LabVIEWTM

251

via RS232 through a TTL-to-RS232 level converter (MAX232, Maxim Integrated 249

Products Inc., Sunnyvale, CA). 250

When MotorLord is executed, the user loads the file generated using WheelScan which contains 252

the raster patterns of the samples to be analyzed. MotorLord then sends digital inhibit signals to 253

the analog-to-digital converter (ADC) and scalar of the RBS/ERD data acquisition system. The 254

software then instructs the vertical motor to move the dynamic stage upward until the 255

microswitch is tripped, establishing the zero reference position for the motor; this procedure 256

recurs between analysis of any separate samples to assure continued precision. As described 257

above, MotorLord also reads the duty cycle of the absolute magnetic encoder which provides the 258

position of the rotary motor. Once the wheel position is properly set to user-chosen locations, 259

the beam current is tuned to the desired value, as needed, by the accelerator operator. Once 260

tuning of the beam current is complete, the user selects the sample to analyze. For a point 261

sample, MotorLord instructs the two motors to move to the corresponding polar coordinates and 262

enables the DAQ, beginning data collection. The user can stop data collection at any time by 263

clicking on a command button. 264

265

For discretely-rastered samples, the user inputs the number of points to be analyzed and the time 266

each point is to be irradiated. MotorLord then instructs first the rotary and then the vertical 267

motors to move to their respective initial sample coordinates. Once at the target coordinate, 268

MotorLord enables data collection. After the specified irradiation time has been reached, 269

MotorLord inhibits data collection, moves the motors so the beam hits the next point on the 270

sample and re-enables data collection. This process continues until all specified points on the 271

sample have been irradiated. 272

273

For continuously-rastered samples, MotorLord prompts the user to input the desired raster speed 274

(mm/s) and analyzes the raster pattern to determine a starting angle 0.1 radians outside the raster 275

area. MotorLord then moves the rotary motor to this starting angle and the vertical motor to the 276

starting radial position of the innermost arc. Next, the software instructs the rotary motor to 277

move to the beginning point of the current arc, enables data collection, and instructs the motor to 278

move at the user-specified speed towards the end angle of the arc. Once there, the rotary motor 279

is returned to the start of the arc and data collection is inhibited before the motor is returned to 280

the starting angle. MotorLord then instructs the vertical motor to move to the radial position of 281

the next arc and the process is repeated until all the arcs have been rastered. 282

283

The user has the option to pause or abort RBS/ERD analysis at any point during a sample run. 284

Once all the desired samples have been analyzed, the user ends execution of MotorLord. 285

286

4. Results 287

288

Figure 3 shows RBS spectra of the ESPA3 reverse osmosis membrane (circles) and its 289

polysulfone support (triangles) (Hydranautics, Oceanside, CA). To facilitate spectra comparison, 290

counts for the polysulfone support have been normalized so that the sulfur plateau matches that 291

of the ESPA3 membrane. The inset in the figure depicts the thin-film composite structure of the 292

ESPA3 membrane and indicates that the active layer is made of polyamide; such structure with a 293

polyamide active layer is the most common structure of reverse osmosis and nanofiltration 294

membranes [2, 6, 10]. The polysulfone support lacks the top polyamide layer and is made up of 295

only the polyester backing and the polysulfone layer. 296

297

Insert Figure 3. Size suggestion: Single column 298

299

The RBS spectra were obtained using the new target system and a 2-MeV 4He2+ analysis beam 300

collimated to a square shape of 3 mm × 3 mm. The membrane samples (i.e., 2.5-cm × 5.0-cm 301

coupons) were attached to the target wheel using double-sided conductive tape, and the ion 302

fluence was always maintained below 3×1014 ions/cm2 to prevent damage of the polymer 303

material [17]. The experiments were performed using incident, exit and scattering angles of 304

22.5o, 42.5o and 160o

310

, respectively, and a sample-detector distance of 75 mm. A collimator with 305

a 6-mm wide rectangular aperture was used in front of the detector. The depth of RBS analysis 306

for the experimental settings used is <5 µm [12] which is much lower than the thickness (~30 307

µm) of the polysulfone layer. As a result, the polyester backing does not contribute to the 308

spectra. 309

While the data in Figure 3 come from experiments performed using RBS geometry and 311

continuously-rastered samples, we also verified the correct performance of the system for ERD 312

geometries, and point and discretely-rastered samples (data not shown). The choice of scanning 313

mode did not affect the data collected, i.e., different scanning modes with the same geometry and 314

fluence generated statistically indistinguishable spectra. Additionally, all scanning modes with 315

RBS geometry resulted in RBS spectra of the membranes analyzed consistent with what is 316

expected from the literature [1-8]; the spectra in Figure 3 are examples of such typical spectra. 317

No ERD spectra have been reported in the literature for RO and NF membranes. The study of 318

such spectra will be the subject of subsequent work. 319

320

The signal in an RBS spectrum is the result of backscattering events of the projectile ions (4He2+) 321

upon collision with atoms in the sample that have a larger mass than the projectile ions [12]. As 322

a result, the spectra in Figure 3 for the ESPA3 polyamide membrane and the polysulfone support 323

contain peaks and plateaus indicative of the presence of only carbon, nitrogen, oxygen, sulfur 324

and chlorine, but not of hydrogen [2, 6]. Also, as depicted in Figure 3, the ESPA3 membrane 325

has a nitrogen peak while the polysulfone support does not. The nitrogen peak is the result of 326

4He2+ scattering from the top polyamide active layer in the ESPA3 membrane [2, 5]. 327

Additionally, since the helium beam loses energy as it travels through the membrane [12] and the 328

top polyamide layer in the ESPA3 membrane does not contain sulfur, the onset of the sulfur 329

signal appears at lower energies in the spectrum of the ESPA3 membrane [2, 5]. Using the 330

theoretical hydrogen-to-carbon ratios of 0.667 and 0.815 for the elemental compositions of 331

polyamide and polysulfone [2, 5, 10, 22], respectively, the commercial software SIMNRA [23] 332

was used to simulate the spectra in Figure 3 (see solid lines). Elemental compositions of 333

C0.50H0.41O0.07S0.02 and C0.49H0.33O0.08N0.08Cl0.01 were obtained for the polysulfone support and 334

the ESPA3 reverse osmosis membrane, respectively. The thickness of the polyamide active 335

layer was estimated at 100 nm assuming a polyamide density of 1.24 g/cm3

338

[7, 21]. All results 336

are consistent with previously published results [7]. 337

Acknowledgements 339

340

The RBS and ERD measurements were carried out in the Triangle Universities Nuclear 341

Laboratories (TUNL), Durham, NC, which is partially supported by the U.S. Department of 342

Energy Office of Nuclear Physics under grants DE-FG02-97ER41041 and DE-FG02-343

97ER41033

348

. The authors gratefully acknowledge Professor Hugon Karwowski and Professor 344

Nalin Parikh for assistance in setting up the scattering system, and John Dunham for assistance 345

in the operation of the tandem accelerator. The company Hydranautics (Oceanside, CA) is also 346

gratefully acknowledged for providing membrane materials. 347

References 349

350

[1] C.R. Bartels, J. Mem. Sci., 45 (1989) 225-245. 351

[2] B. Mi, O. Coronell, B.J. Mariñas, F. Watanabe, D.G. Cahill, I. Petrov, J. Mem. Sci., 282 352

(2006) 71-81. 353

[3] B. Mi, B.J. Mariñas, D.G. Cahill, Environ. Sci. Technol., 41 (2007) 3290-3295. 354

[4] X. Zhang, D.G. Cahill, O. Coronell, B.J. Mariñas, Appl. Phys. Lett., 91 (2007) 181904. 355

[5] O. Coronell, B.J. Mariñas, X. Zhang, D.G. Cahill, Environ. Sci. Technol., 42 (2008) 5260-356

5266. 357

[6] O. Coronell, B.J. Mariñas, D.G. Cahill, Environ. Sci. Technol., 43 (2009) 5042-5048. 358

[7] O. Coronell, M.I. González, B.J. Mariñas, D.G. Cahill, Environ. Sci. Technol., 44 (2010) 359

6808-6814. 360

[8] O. Coronell, B.J. Mariñas, D.G. Cahill, Environ. Sci. Technol., 45 (2011) 4513-4520. 361

[9] T. Maugin, T. Renkens, A. Martínez Sáenz de Jubera, J.P. Croué, B.J. Mariñas, S. Rapenne, 362

Fate of reverse osmosis (RO) membranes during desalination of chloraminated seawater, in: 363

Proceedings of the American Water Works Association Membrane Technology Conference, 364

Long Beach, CA, 2011. 365

[10] R.J. Petersen, J. Mem. Sci., 83 (1993) 81-150. 366

[11] M. Mulder, Principles of Membrane Technology, Second ed., Kluwer Academic, Boston, 367

MA, 1996. 368

[12] W.-K. Chu, J.W. Mayer, M.A. Nicolet, Backscattering spectrometry, Academic Press, New 369

York, NY, 1978. 370

[13] F. Namavar, J.I. Budnick, Nucl. Instrum. Meth. B, B15 (1986) 285-287. 371

[14] T. Venkatesan, W.L. Brown, C.A. Murray, K.J. Marcantonio, B.J. Wilkens, Polym. Eng. 372

Sci., 23 (1983) 931-934. 373

[15] A.L. Evelyn, D. Ila, R.L. Zimmerman, K. Bhat, D.B. Poker, D.K. Hensley, C. Klatt, S. 374

Kalbitzer, N. Just, C. Drevet, Nucl. Instrum. Meth. B, B148 (1999) 1141-1145. 375

[16] L.J. Matienzo, F. Emmi, D.C. Van Hart, T.P. Gall, J. Vac. Sci. Technol. A, 7 (1989) 1784-376

1789. 377

[17] B. Mi, D.G. Cahill, B.J. Mariñas, J. Mem. Sci., 291 (2007) 77-85. 378

[18] S. Bason, V. Freger, J. Mem. Sci., 302 (2007) 1-9. 379

[19] V. Freger, Environ. Sci. Technol., 38 (2004) 3168-3175. 380

[20] F.A. Pacheco, I. Pinnau, M. Reinhard, J.O. Leckie, J. Mem. Sci., 358 (2010) 51-59. 381

[21] X. Zhang, D.G. Cahill, O. Coronell, B.J. Mariñas, J. Mem. Sci., 331 (2009) 143-151. 382

[22] K. Ikeda, S. Yamamoto, H. Ito Sulfonated polysulfone composite semipermeable 383

membranes. European Patent, EP 0 165 077 B2, 1998. 384

[23] M. Mayer, J.L. Duggan, I.L. Morgan, SIMNRA, a simulation program for the analysis of 385

NRA, RBS and ERDA, in: Proceedings of the 15th International Conference on the Application 386

of Accelerators in Research and Industry, Denton, TX, 1998. 387

388

389

Set Reference Position

RBS/ERD Data Acquisition

System

Encoder Signal

Target Wheel Azimuthal Position

Raster Pattern

Enable/Inhibit Signal Target Wheel

Vertical Position

Rotary Motor

PIC 18F4520 Microcontroller

User Input: Raster Speed and Sample

Number

Vertical Motor

TMCM 310/SG Motor Driver/Controller

Micro- switch

LabView VI MotorLord Data

LabView VI WheelScan

User Input: Digital Image

of Target Wheel

User Input: Beam Width, Beam Angle, Number of

Samples, Type of Sample, Sample

Boundaries