Improved Separation of Pa from Th and U in Marine Sediments with

Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES

Article

Improved Separation of Pa from Th and U in marine Sediments with TK400 Resin Finn Süfke, Jörg Lippold, and Steffen Happel Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b04723 • Publication Date (Web): 19 Dec 2017 Downloaded from http://pubs.acs.org on December 20, 2017

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

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

Page 1 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Improved Separation of Pa from Th and U in marine Sediments with TK400 Resin

1 2 3

Finn Süfke†*, Jörg Lippold†, and Steffen Happel‡

4 5 6

†

7

*Corresponding Author: Finn Süfke ([email protected])

Institute of Earth Science, Heidelberg University, Im Neuenheimer Feld 234, 69120 Heidelberg, Germany TrisKem International, 3 rue des champs Geons, 35170 Bruz, France

‡

8 9

Abstract

10

Protactinium-231 is a radio-nuclide of broad interest in paleoceanography and

11

paleoclimatology. This study describes an improved method for the purification and

12

separation of Pa from marine sediment samples using the new TK400 resin by TrisKem

13

International. The focus lies on the improvement of the separation of the abundant

14

from the Pa-fraction of the sample, which would reduce the peak tailing from

15

masses 231 and 233 during ICP-MS measurement. Furthermore, the reusability of TK400 has

16

been tested. For this purpose, the conventional method using Dowex AG1X8 for the

17

separation and purification of Pa has been compared to methods using the TK400 resin. A

18

combination of a Dowex AG1X8 prior to a TK400 column has shown most convincing results.

19

Based on our results we suggest a new efficient procedural method of analyzing 231Pa from

20

marine sediment samples using TK400. Chemical Pa yields for a Dowex-TK400 combination

21

are highest compared to the Dowex only method. Furthermore, the 232Th/231Pa ratio of the

22

Pa-fractions has been reduced by one order of magnitude compared to conventional

23

methods with Dowex AG1X8. Additionally, the reusability of TK400 resin up to nine times has

24

been proven. The usage of TK400 is only limited in the presence of samples with a high

25

matrix load (e.g. Fe). Therefore, matrix from sediment samples needs to be removed (here

26

using Dowex resin) before samples are loaded onto TK400. We also report series of

27

concentration measurements from standard reference materials (UREM-11, NIST 2702),

28

which have been used for 233Pa calibration.

29 30

Introduction

ACS Paragon Plus Environment

232

Th

232

Th on

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 24

31

With the improvement of measuring techniques, low concentrated trace-elements and their

32

isotopes have received increasing interest in ocean science. One set of these isotopes are

33

231

34

sedimentation and accumulation rates,1,2 paleoproductivity and boundary scavenging3,4,5 as

35

well as a proxy for ocean circulation strength (231Pa/230Th)6,7,8. Earliest studies were

36

performed with α and β spectrometry. While the inductively coupled plasma mass

37

spectrometry (ICP-MS) has emerged as the method of choice nowadays there also have

38

been successful attempts by TIMS9 and even AMS10.

39

Today ICP-MS is a widespread tool for the determination of isotopic composition. With its

40

high ionization efficiency this measuring technique allows the quantification of low

41

abundances.11 The concentration of 231Pa in marine sediments is typical in the order of very

42

few pg/g and decreases with age due to the relative short half-life of 32.5 ka12. But besides

43

the measurement method the preceding chemical preparation and separation of

44

bearing samples is essential to guarantee reliable results.

45

In early studies13,14 Pa and Th have been separated from marine sediment samples using the

46

strong anion exchange resin Dowex AG1X8. Until today this resin represents the most

47

common agent for this purpose.15 With the development of new resins separation methods

48

have been improved and modified. In addition to Dowex other commonly used resins are

49

TRU,16,17 Silica gel,18 TEVA19 and UTEVA20. Separation methods have been modified for

50

specific applications like sediment samples,20,21 seawater samples22,23 and nuclear material18.

Pa and

230

Th. In oceanography and paleoceanography they are applied as proxies for

231

231

Pa

Pa-

51 52

In particular, the chemical separation of Pa from marine sediments should fulfill following

53

requirements (a) good removal of sample matrix (b) separation of Pa from Th to prevent

54

peak tailing effects of

55

separation of Pa from U to minimize the influence of the

56

mother nuclide (d) and the chemical procedure should also be rapid and low-cost.

57

Today’s method of accessing Pa and Th from marine sediments includes the total dissolution

58

of the sample material. Consequently, processed sediment solutions generally hold a very

59

high matrix load (dissolved ions). Matrix consists primarily of major elements like e.g. Si, Fe,

60

Ca and Al which cause problems during separation and measuring of Pa due to their sheer

61

abundance. During separation a high matrix load can occupy ion-binding sites in resins and

62

prevent target ions like e.g. Pa from retention on the resin. Furthermore, if the matrix is not

232

Th on masses 231 and 233 during ICP-MS measurement (c)


233

Pa decay product

233

U on its

Page 3 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


63

removed before measurement, the large ion load can lead to major interferences when

64

using mass spectrometers.

65

Also a good separation between Th and Pa is essential for reliable measurements of the

66

231

67

measured using the ion counting mode of the ICP-MS at low resolution. Accordingly, both Pa

68

isotopes can be easily overprinted by a high 232Th peak tailing on masses 231 and 233 (Figure

69

1). The average

70

optimal separation of both elements is the foremost concern of every column

71

chromatography approach.

72

Most resins used today are expensive and thus the amount of potentially processed samples

73

is limited. Furthermore, the quality of separation is linked to the amount of resin and acids,

74

and the use of a large resin bed can result in a very time consuming procedure. In science

75

cost and time should not be the limiting factor, but a proper method for daily use should be

76

in a reasonable relationship to costs and time.

77

In this study a modified separation method for Th, U and Pa is tested. Here the established

78

Dowex AG1X8 resin is compared to the new TK400 resin (TrisKem) which comprised 1-

79

octanol adsorbed to a semi-porous resin material24. The TK400 resin is expected to provide a

80

good separation quality for Th and Pa25 but has not been tested with marine sediment

81

samples yet. Here we assess the quality of the Pa separation from U and Th in marine

82

sediments as well as the reusability of the TK400 resin as a factor for reducing costs.

83

Experimental Section

84

Safety Considerations

85

For microwave digestion of sediment samples 48 % HF has been used. HF should only be

86

handled following the compulsory laboratory rules in a fume hood connected to a washing

87

system for acid fumes. Radioactive

88

emitter and should only be handled in appropriate laboratories designed and accredited for

89

radioactive substances.

90

Standard Sample Solution

91

All of the following described experiments have been performed using the same marine

92

sample material in order to guarantee unbiased comparability of the tested methods. For

Pa/233Pa ratio. Both Pa isotopes are present in very low concentrations and have to be

232

Th/231Pa ratio in marine sediments is in the range of ~ 2x106. Thus, an

237

Np used for the preparation of the


233

Pa spike is a α-


93

this purpose a composite of several deep sea sediment samples (~4 g in total) has been

94

dissolved by microwave digestion and homogenized representing an average matrix solution

95

typical for average deep sea marine sediments. This standard solution was enriched by few

96

ml of a 231Pa-solution obtained from sample remains from preceding 231Pa measurements in

97

order to achieve reasonable high 231Pa levels. By this 250 ml of standard solution have been

98

produced. 5 ml of this solution corresponds to ~100 mg of dried and mortared sediment and

99

represents one sample for the following described method analyses. All samples and

100

aliquots have exclusively been handled in PFA vials to prevent Pa from adhering to vial walls.

101

Experiments

102

The TK400 resin has been tested with the marine sediment standard solution in three

103

experimental sets. (1) TK400 resin has been directly compared to the conventional

104

established separation method using Dowex AG1X8. (2) The reusability of already used

105

TK400 resin has been tested. (3) Particle size of the TK400 resin and column geometry have

106

been varied. Every test has been performed with at least two duplicates in order to ensure

107

reliability of the results.

108

(1) TK400 (100-150 µm) vs. Dowex AG1X8

109

In a first experiment two subsequent applied Dowex AG1X8 columns following a method

110

slightly modified from published procedures21,14,26 have been compared to a method using

111

one single TK400 column applying a slightly modified procedure suggested by Jerome et al.

112

(2017)25 and Knight et al. (2016)24. Finally, a combination using one Dowex AG1X8 column

113

followed by one column of TK400 applying the above mentioned methods has been tested.

114 115

For the first Dowex column a Bio-Rad Poly-Prep column is filled with 5 ml Dowex AG1X8. The

116

resin is cleaned using 15 ml H2O (MQ-quality) and conditioned with 25 ml 8M HNO3. After

117

loading the sample solution (5 ml) onto the column and elution of the matrix using 20 ml 8M

118

HNO3, Th is eluted using 20 ml 9M HCl and 1 ml 9M HCl + 0.13M HF (4:1). Afterwards Pa is

119

eluted with 10 ml 9M HCl + 0.13M HF (4:1). Finally, U can be eluted with H2O. The Pa fraction

120

is cooked to nearly dryness in concentrated HNO3 with drops of boracic acid to remove

121

fluorides and then transferred into chloride form with concentrated HCl. Afterwards the Pa

122

fraction is taken up in 9M HCl for the second Pa purification column. The column is filled

123

with 3 ml Dowex AG1X8 and conditioned with 15 ml 9M HCl. After loading the sample, Th is ACS Paragon Plus Environment

Page 4 of 24

Page 5 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


124

removed with additional 15 ml of 9M HCl. Pa is eluted with 10 ml 9M HCl + 0.13M HF (4:1).

125 126

For the test using TK400 a modification of published protocols24,25 has been introduced. 2 ml

127

TrisKem columns (AC-142-TK) have been filled with 2 ml TK400 (100-150 µm). The resin has

128

been cleaned with one column volume (cv) of 1M HCl and one cv H2O (MQ-quality).

129

Afterwards the resin has been conditioned with one cv 10M HCl before the sample has been

130

loaded onto the resin. 20 ml of 10M HCl have been used to remove Th. Pa has been eluted

131

with 20 ml of 1M HCl. After the elution of Pa the resin has been cleaned with two cv 1M HCl

132

and one cv H2O (MQ-quality). Finally, the column including the resin has been stored in H2O

133

for future use. This separation protocol does not allow the individual separation of a Th, U

134

and Pa fraction and was tested to evaluate the ability of TK400 separating Pa directly from a

135

sediment solution.

136 137

For this reason a combination of first applying a Dowex column followed by a TK400 column

138

for the purification of the Pa fraction has been tested.

139

The first 10 ml of the Pa fraction from TK400 of all these experiments have been divided by

140

collecting the eluate in 2.5 ml steps to monitor the elution of the columns. The last 10 ml of

141

the Pa fraction have been collected in 5 ml steps since Pa was expected to be present at low

142

concentrations only here.

143

(2) Reusability of TK400 (100-150 µm)

144

In a second experiment the reusability of TK400 has been tested. For this purpose three

145

individual columns have been used repetitively. One column has been run just with acid

146

without sample material. This blank-column was used to monitor changes in the resin due to

147

stress from the recurring load with strong acid (10M HCl) and to monitor eventual

148

contaminations (Th and Pa) from the used chemicals or handling. The other two columns

149

have been used with the standard sample solution three times in a row. Afterwards a blank

150

run without sample solution has been performed to monitor an eventual memory effect of

151

Pa in the resin. This procedure has been repeated three times. Overall TK400 has been

152

tested nine times with the standard sample solution on two columns. For this experiment

153

the afore described procedure has been applied (samples have been run first through a

154

Dowex column before being loaded on TK400) with the exception that only 10 ml of the Pa

155

fraction have been collected in 5 ml steps. ACS Paragon Plus Environment


Page 6 of 24

156

(3) Change in particle size, resin volume and column geometry

157

In a third experimental set the procedures of the first experiment have been repeated but

158

with modified TK400 column volumes or smaller particle sizes of the resin. The resin volume

159

has been reduced to 1 ml to monitor if results from the first experiments can also be

160

reproduced in shorter time and at lower costs. Further the geometry of the column has been

161

changed to a wider column diameter. Finally, 2 ml of TK400 with a smaller particle size (50-

162

100 µm) has been tested. This experiment was expected to be more time consuming but

163

may result in a better separation of Th and Pa as well as in higher chemical Pa yield.

164

233

165

To monitor the

166

from 237Np using silica gel columns11. For the determination of the original 231Pa content of

167

the standard solution a few samples have been spiked before column chemistry. All other

168

samples have been spiked after column chemistry in order to define the chemical yield of

169

the various chemical treatments. In theory uncertainties in the yield calculation may arise

170

from the ingrowth of

171

treatment as no chemical separation of 233U and 233Pa has been applied. However, because

172

measurements have been performed within less than one week after chemical procedures

173

and because of all experiments have been performed with the identical spike solution all

174

results are highly comparable and distortions from 233U are thus negligible.

175

The 233Pa spike was calibrated against an in-house pitchblende standard28. The reliability and

176

the trueness of the obtained concentrations have been monitored by co-measuring of the

177

UREM-11 standard material29. According to Hansen and Ring (1983)30 UREM-11 is in

178

radioactive equilibrium featuring a certified uranium concentration of 58.48 ppm. From this

179

value a

180

repeatedly

181

term mean obtained in the years 2009 to 2017 is 34.00 ± 0.36 Bq/kg in reasonable

182

agreement to the expected value (Figure 2). However, two disadvantages of UREM-11 as a

183

standard material for measuring 231Pa/230Th from marine sediments becomes obvious: First,

184

its

185

Figure 2) and close only to very young and deep sediments from the opal rich Southern

186

Ocean31,32,3. Due to its high concentration very little net weight (max. 50 mg) of UREM-11

Pa-Spike calibration and Yield Calculation

231

231

231

Pa recovery, samples have been spiked with 233Pa.

233

U from

233

Pa has been milked

233

Pa (t1/2: 26.975 d)27 for samples spiked after chemical

Pa concentration of 33.5 ± 0.3 Bq/kg can be deduced29,20. The results of the 231

Pa-measurements of UREM-11 in Heidelberg is shown in Figure 2. The long-

Pa-concentration is way higher than that of typical deep-sea sediment (see green bar


Page 7 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


231

187

has been used per sample. This, however, may lead to variations in the measured

188

concentration due to faint but obvious inhomogeneous distributions of

189

material. When using higher quantities of UREM-11 (e.g. >100 mg) these inhomogeneities

190

are tend to be averaged out. This behavior is also recognizable for 232Th and 230Th values of

191

UREM-11. That’s why other marine standard material might be more appropriate in order to

192

monitor the quality of the obtained isotopic concentrations from marine sediments. The

193

most comprehensive attempt so far providing a marine-based

194

been performed by an international inter-lab-calibration in the framework of GEOTRACES

195

distributing siliceous ooze sediment to a large number of laboratories23. The average

196

concentration of this material among the contributing laboratories yielded 20.83 ± 2.33

197

Bq/kg in good agreement to subsequent measurements at Heidelberg University (21.63 ±

198

0.23 Bq/kg, Figure 2). Other established sediment reference materials are rarely certified

199

and/or intercalibrated for 231Pa concentrations. For guidance here we provide values of 231Pa

200

concentrations for the commonly used NIST SRM 2702 measured in our lab within the last

201

years (Figure 2)33,34.

202

A third disadvantage of UREM-11 as a standard material for measuring

203

excess (that is the fraction of both radioisotopes produced in-situ from the decay of 235U and

204

234

231

U) from marine sediments is given by the lack of a certified 232

Pa-

231

Pa within the

Pa-reference material has

232

231

Pa- and

231

230

Pa

Th-

Th-concentration for this

205

standard material. Knowledge about the

206

sample is required in order to correct for lithogenic contributions35,36. Here we report a value

207

based on 55 individual repetitive measurements of the UREM-11

208

yielding an average of 42.87 ± 0.34 Bq/kg (supplementary information).

209

Measurement

210

Radionuclides

211

Element2 single collector ICP-MS at the Institute of Earth Sciences, Heidelberg University,

212

Germany with an APEX IR desolvating system. To prevent clogging all samples have been

213

filtered before being placed in the auto-sampler. All isotopes were measured in low

214

resolution mode by ion counting. Washing steps after each sample comprised a 2M nitric

215

acid solution doped with traces of HF to efficiently remove Th and Pa from hoses and reduce

216

memory effects (washs). The washs have been used for background corrections, but never

217

reached more than 0.5 % of average sample signal on

218

influence of

231

232

Pa,

232

Th,

233

Pa and

236

Th concentration of the examined sediment

232

Th concentration,

U have been measured using a Thermo Finnigan

231

Pa. In order to monitor the

Th peak tailing on masses 231 and 233 a Th-standard solution without ACS Paragon Plus Environment


231

Pa and 233U/233Pa has been measured in intervals. Instrumental drift has

219

contributions of

220

been monitored by using an in-house synthetic 231Pa-233U solution.

221 222

Results and Discussion

223

Chemical Yield and Purity

224

The standard sample solution had a

225

concentrations of the collected

226

0.052 pg/g ± 0.001 for Dowex-columns to 0.087 pg/g ± 0.002 for a Dowex-TK400

227

combination (Figure 3a).

228

Accordingly, chemical Pa yield is 27 ± 4 % for two successive Dowex columns implying that a

229

single Dowex column recovers 9M) HCl. Indeed, our TK400 (100-150 µm)

255

tests yielded

256

eluted with 1M HCl. Results for the first eluted 5 ml are clearly better with 232Th/231Pa ratios

257

even below 30. Since this first 5 ml contains the majority of Pa, elution is recommended to

258

be stopped at this point if an exceptionally clean Pa fraction is wanted, however at the

259

expense of slightly lower recovery.

260

Furthermore, experiments with less resin or finer particle sized (50-100 µm) TK400 have

261

obtained slightly worse results with 232Th/231Pa ratios ranging between 100 and 600, but still

262

better than separation with Dowex. Accordingly, summarizing all of the here presented

263

results the method applying 2 ml TK400 with a particle size of 100-150 µm seems the most

264

effective procedure for separating Pa from Th.

265

In order to monitor the separation of Pa from U, samples have been spiked after the first

266

Dowex column with a

267

respectively. The initial

268

reduced this ratio down to 236U/231Pa ~4-6 in the first collected 10 ml. Therefore, Dowex and

269

TK400 appear to be equally efficient in the separation of U and Pa.

270

Reusability

271

In a further experiment the reusability of TK400 resin (100-150 µm) has been tested. The

272

limit of reuse is monitored by the recovered amount of 231Pa in the collected Pa fraction as

273

well as the 232Th/231Pa ratio. The Pa fraction has been collected in two 5 ml steps. The first 5

274

ml of the Pa fractions, which hold the predominant portion of the eluted Pa as seen from

275

prior experiments, have been spiked with

276

are on a very constant level (average 0.061 ± 0.004 pg/g, Figure 5). Furthermore, even after

277

the ninefold use of the same TK400 resin chemical yields are still clearly better than found

278

for the Dowex column used only once. This indication for good reusable is also confirmed by

279

the 232Th/231Pa ratios repeatedly below 80 for the first 5 ml of the collected Pa fraction and

232

Th/231Pa ratios always below 200 for the collected 10 ml of the Pa-fraction

236 236

U spike for both TK400 and Dowex as the second column

U/231Pa ratio before the second column was ~4,400. Both resins

233

Pa.

231

Pa concentrations of the spiked aliquots



Page 10 of 24

280

below 120 for the complete Pa fractions (Figure 5). These results are in the same order as

281

seen in the first experiments when testing TK400 (100-150 µm) versus Dowex performance.

282

All these results suggest that TK400 (100-150 µm) is a robust resin that shows no loss of

283

performance even after multiple use. Although notably the column flow through extended

284

by one hour from the first to the ninth repetition with the standard sample solution.

285

The cleaning procedure has been monitored by collecting and measuring blank solutions

286

rinsed through the cleaned resin. All Blanks did show no measurable accumulation of Pa in

287

the TK400 resin. Counts on 231Pa and 233Pa in the blanks have been negligibly low and can be

288

interpreted as artefacts (

Improved Separation of Pa from Th and U in Marine Sediments with

Recommend Documents