Follow-Up Study of the Humic Substances and Colloidal Material in Groundwater from ONKALO,

Olkiluoto, 2017

Wo

rk

ing

Re

po

rt 2

01

7-0

2 •

Ge

oc

he

mic

al a

nd

Ph

ys

ica

l Pro

pe

rties

an

d in

Situ

Dis

tribu

tion

Co

effic

ien

ts o

f the

De

ep

So

il Pits

OL-K

K25 A

ND

OL-K

K26 a

t Olk

iluoto

POSIVA OY

Olki luoto

FI-27160 EURAJOKI, F INLAND

Phone (02) 8372 31 (nat. ) , (+358-2-) 8372 31 ( int. )

Fax (02) 8372 3809 (nat. ) , (+358-2-) 8372 3809 ( int. )

June 2019

Working Report 2019-08

Sari Luste

June 2019

ONKALO is a registered trademark of Posiva Oy

Working Reports contain information on work in progress

or pending completion.

Sari Luste

Eurof ins Ahma Oy

Working Report 2019-08

Follow-Up Study of the Humic Substances and Colloidal Material in Groundwater from ONKALO,

Olkiluoto, 2017

FOLLOW-UP STUDY OF THE HUMIC SUBSTANCES AND COLLOIDAL MATERIAL IN GROUNDWATER FROM ONKALO, OLKILUOTO, 2017

ABSTRACT The humic substances (HS) at groundwater from ONKALO®, Olkiluoto were studied in order to determine the apparent molecular size distribution and the amount of humic substances and to compare the results with the previous years’ results. Groundwater samples were taken from the groundwater stations ONK-PVA3, ONK-PVA5 and ONK-PVA10. Humic substances were isolated and molecular size distribution was determined using high resolution size exclusion chromatography (HPSEC) method optimized during previous years. The amount of humic substances was determined by total organic carbon analyser (TOC). In addition, metal content of initial and isolated humic fraction was studied. The results of both molecular size distribution and amount of humic substances were similar with previous studies.

In addition to HS also a study related to colloids was performed. The purpose of this study was to estimate the concentration of groundwater colloids and to compare the results with the previous ones. The water samples were collected from groundwater stations ONK-PVA5, ONK-PVA10 and ONK-PVA11.

The colloid concentrations were determined from scanning electron microscopy (SEM) micrographs taken from the filters in which the groundwater was run through. The particle calculations were performed with computer software ImageJ. The colloid concentration (Ø 0–1 µm) determined by the single particle analysis of SEM micrographs was somewhat higher in ONK-PVA5 and ONK-PVA10 compared to previous study. Colloids could not be evaluated from ONK-PVA11 since the filter used for sampling was blocked due to long sampling time.

Keywords: DAX-8, colloids, fulvic acids, groundwater, humic acids, HPSEC, SEM

HUMUSTEN JA KOLLOIDISTEN PARTIKKELIEN SEURANTATUTKIMUS OLKILUODON ONKALOSSA 2017

TIIVISTELMÄ

Olkiluodon ONKALOn pohjaveden humusyhdisteiden molekyylikokojakaumaa ja määrää tutkittiin ja tuloksia vertailtiin aiempien vuosien tutkimustuloksiin. Pohjavesinäytteet otettiin asemilta ONK-PVA3, ONK-PV5 ja ONK-PVA10. Humusyhdisteet eristettiin ja molekyylikokojakaumaa määritettiin aiempina vuosina optimoidulla korkean erotuskyvyn kokoekskluusiokromatografisella (HPSEC) menetelmällä. Humusten määrä laskettiin kokonaisorgaanisen hiilen (TOC) määritysmenetelmää käyttäen. Lisäksi näytteiden metallipitoisuuksia tutkittiin. Sekä humusten molekyylikokojakauma että määrä olivat aiempina vuosina saatujen tulosten kanssa yhteneviä.

2017 tutkittiin myös pohjaveden kolloideja erillisessä tutkimuksessa. Työn tarkoitus oli arvioida pohjavedessä esiintyvien kolloidien määrää ja verrata tuloksia aiempiin tuloksiin. Näytteet haettiin pohjavesiasemilta ONK-PVA5, ONK-PVA10 ja ONK-PVA11. Kolloidien pitoisuus määritettiin pyyhkäisyelektronimikroskoopilla otetuista mikrografeista suodattimilta, joiden läpi pohjavettä laskettiin. Partikkelilaskenta suoritettiin ohjelmaa ImageJ apuna käyttäen. Kolloidien konsentraatio (halkaisija 0–1 µm) SEM kuvista tehtyjen laskentojen perusteella pisteissä ONK-PVA5 ja ONK-PVA10 jonkinverran aiempaa tutkimusta korkeammalla tasolla. Pisteestä ONK-PVA11 ei kolloideja saatu määritettyä suodattimien tukkeutuessa pitkästä näytteenottoajasta johtuen.

Avainsanat: DAX-8, fulvohapot, pohjavesi, humushapot, HPSEC, kolloidit, SEM

1

TABLE OF CONTENTS

ABSTRACT

TIIVISTELMÄ

1. INTRODUCTION................................................................................................... 5

1.1 Humic substances ......................................................................................... 5 1.2 Colloidal material ........................................................................................... 6

2. HUMIC SUBSTANCES ......................................................................................... 7

2.1 Sampling ....................................................................................................... 7 2.2 Chemicals and standards .............................................................................. 7 2.3 Sample isolation ............................................................................................ 7 2.4 TOC analysis ................................................................................................. 8 2.5 Molecular size distribution ............................................................................. 8

2.5.1 Analytical method ...................................................................................... 8 2.5.2 Equations for calculatio .............................................................................. 9

2.6 ICP-MS/MS analysis method ......................................................................... 9 2.7 Results and discussion ................................................................................ 10

2.7.1 Amount of HS .......................................................................................... 10 2.7.2 Molecular size distribution ....................................................................... 11 2.7.3 Analysis of metal content in groundwater samples by ICP-MS/MS .......... 13

2.8 Conclusions ................................................................................................. 15 3. COLLOIDS .......................................................................................................... 17

3.1 Sampling preparations and sampling procedures ........................................ 17 3.2 Single particle analysis method ................................................................... 17 3.3 Results ........................................................................................................ 19 3.4 Conclusions ................................................................................................. 24

RECOMMENDATIONS ............................................................................................... 25

REFERENCES ........................................................................................................... 27

2

3

LIST OF SYMBOLS AND ABBREVIATIONS ρ Polydispersivity ACN Acetonitrile DOC Dissolved organic carbon DOM Dissolved organic matter FA Fulvic acids HA Humic acids HPSEC High performance size exclusion chromatography HS Humic substances IHSS International Humic Substances Society LOQ Limit of quantification mDa Millidalton Mn Number-average molecular weight Mw Weight-average molecular weight MS Mass spectrometry MW Molecular weight NOM Natural organic matter PSS Poly(styrene sulfonate) sodium salt QC Quality control RI Refractive index RT Retention time SEC Size exclusion chromatography SEM Scanning electron microscope TOC Total organic carbon UPLC Ultra performance liquid chromatography UV Ultraviolet UV254 Ultraviolet detection at 254 nm ICP-MS/MS Inductively coupled plasma tandem mass spectrometer

4

5

1. INTRODUCTION

ONKALO is the underground rock characterisation facility that has been constructed by Posiva for spent nuclear fuel disposal research purposes. For long-term safety evaluation, characterisation of groundwater aquatic humic substances (HS) and the overall colloidal material is essential. Colloids in natural groundwaters consist of inorganic material (e.g. clay, oxides and carbonates) and organic material (mainly humic and fulvic acids). The approach is totally different when studying HS or colloids and thus the report is devided into two independent sections.

1.1 Humic substances

Aquatic HS are major components of the natural organic matter (NOM) in soil and water. They generally comprise one third to over half of the dissolved organic carbon (DOC) in natural surface waters. HS are complex organic compounds, which are mainly derived from soil humus and terrestrial and aquatic plants. The colour of the dissolved and precipitated HS varies from light to dark brown. The precise composition and properties of the HS varies depending on the source. HS are highly chemically reactive but recalcitrant to biodegradation and among dissolved organic matter (DOM) have a significant role in the physical, chemical and microbiological processes in water ecosystems. For instance they have an effect on the transportation, toxicity and bioavailability of metals. Because of the low concentration of aquatic HS, the material must first be isolated and concentrated for further studies. The fraction of aquatic HS, that is not soluble in water at a pH lower than 2, are defined as humic acids (HA). The fraction of HS that is soluble under all pH conditions are referred to as fulvic acids (FA) (Malcolm 1990). The method used in concentrating and isolating the dissolved organic matter should be effective in isolation of the aquatic humic matter at the same time keeping the chemical and structural composition unaffected as it occurs in its original state in an aquatic environment (Aiken et al. 1985). The most frequently used isolation procedure for HS is the column chromatographic method by non-ionic sorbents such as XAD resins (polymethyl methacrylate) or analogous materials (Leenheer 1981, Thurman et al. 1981, Peuravuori et al. 1997).

Molecular weight (MW) is one of the fundamental properties that need to be known in order to understand the physical and chemical characteristics and chemical reactivity of HS. The MW of HS has an effect especially on their adsorption and metal complexing behaviour. Determination of the MWs is a difficult task because HS are a complex mixture of natural, heterogeneous organic materials with different structures and a broad molecular size distribution. Size exclusion chromatography (SEC) has been frequently used especially for characterisation of aquatic and soil HS (Janoš and Zatřepálková 2007).

This study continues previous work among HS characterisation. Detailed information of the geo-chemical behaviour of radioactive and thus possibly hazardous metal ions under environmental conditions is necessary for the long-term storage of radioactive waste. In the immobilisation or mobilisation of the metal ions due to the complexation and colloid formation, HS and other DOM fractions might play an important role. The goal of the current study was to determine the amount and the molecular size distribution of HS in the groundwater in the bedrock at Olkiluoto in 2017 and to compare the results with the

6

previous year’s data. Since the molecular weights are strongly operational, the values calculated from the SEC measurements are "apparent molecular weights".

1.2 Colloidal material

Characterisation of groundwater colloids is essential owing to the fact that groundwater colloids may transport radionuclides released from a failed spent fuel canister. Colloids are generally referred to as particles that do not appear to be sedimented even after a long period of time. The sizes of colloid particles are usually from 1 nm to 1 000 nm. Owing to their small size, nano-micrometer particles may be transported along fractures in the bedrock. The transportation is controlled by groundwater movement, attachment-detachment kinetics and sedimentation rates. Colloid formation and stability is dependent on the surrounding hydrogeochemical conditions e.g. pH, conductivity and redox. Generally colloids tend to form agglomerates in saline conditions and therefore the colloid concentrations are lower in saline waters than in fresh waters.

The colloids of interest are colloids that reflect the in-situ conditions. During the sampling the in-situ conditions are tried to maintain at the sampling locations by avoiding sudden changes in flow rates and keeping the flow rates steady.

The colloid concentrations were determined by single particle counting of scanning electron microscopy (SEM) micrographs taken from the filters. The particle calculations and size distribution determinations were performed with the assistance of ImageJ, an image analysis program.

The purpose of this study was to analyse the colloid content and to compare the results with the previous ones. Enhancements in the SEM method were expected to increase the reliability of the results.

The groundwater stations in ONKALO are presented in figure 1.

Figure 1. The groundwater monitoring stations in ONKALO, Olkiluoto (Pöyry Finland Oy).

7

2. HUMIC SUBSTANCES

2.1 Sampling

Groundwater samples were collected on the 23th of March 2017 from groundwater stations ONK-PVA3 (depth -85 metres above sea level (masl)), ONK-PVA5 (depth -229 masl) and ONK-PVA10 (depth -366 masl) in ONKALO, Olkiluoto, Finland. Figure 1 illustrates the locations of the groundwater stations in ONKALO.

The samples were on-line pressure filtered in situ, using Whatman mixed cellulose ester, 0.45 μm filters. The tubing material used in the sampling was nylon. 2 l glass bottles were used to collect and store the samples. Three parallel 2 l samples were taken from every sampling point. The samples were stored at 6 ºC and, since the photosensitivity of HS (Vidali et al. 2010), protected from the light by wrapping the bottles in aluminium foil.

2.2 Chemicals and standards

All the solutions used were made using high-purity MQ water.

Methanol (J. T. Baker) was of the “Baker Analysed” quality.

n-Hexane was of the “Baker resi-analysed” quality (J. T. Baker).

Analysis grade 1.0 M HCl and NaOH (Reagecon).

Acetonitrile (Prolabo) was of LC-MS grade.

Phosphate buffer was made of K2HPO4 and KH2PO4, both "EMSURE®" (99%) quality (Merck).

DAX-8 resin (Supelite DAX-8, Supelco) was used for the isolation of HS from groundwater.

Reference solution containing humic and fulvic acids were made by dissolving International Humic Substances Society (IHSS) Nordic Aquatic Humic Acid Reference (1R105H) and IHSS Nordic Aquatic Fulvic Acid Reference (1R105F) in phosphate buffer. The reference humic and fulvic acids were dissolved by shaking.

Poly(styrene sulfonate) sodium salt standards (PSS) with MWs 210, 3 610 and 6 520 g/mol were used as a calibration standards for the molecular size distribution investigations. PSS standards were received from PSS Polymer Standards Service GmbH, except for the 210 g/mol standard, which was ordered from Sigma-Aldrich.

QCWW4 reference solution (Eurofins A/S) was used as a quality control sample for the total organic carbon (TOC) determinations.

2.3 Sample isolation

Prior to isolation, DAX-8 was first cleaned according to Thurman et al. (1981) with slight modifications. The DAX-8 was cleaned by rinsing the resin in a beaker with 0.1 M NaOH and thereafter with methanol until pH were neutral. Next, the resin was Soxhlet-extracted sequentially for 6 h with n-hexane and methanol. The resin was stored in methanol at 6 ºC until used.

8

Isolation of the organic fractions were made similarly to previous studies with humic and fulvic acids (Luste et al. 2012, Luste et al. 2013). For the isolation of HS the cleaned DAX-8 polymer in methanol was slurry-packed into a 160-mm I.D. class column, resulting in an 8-cm sorbent bed. The packed column was rinsed with MQ water until it was free of methanol. The groundwater sample, acidified to a pH 2.5 with 1.0 M HCl, was placed in a separatory funnel and slowly drained through the resin. The humic substances (HS) adsorbed on the resin were eluted with 60 ml of 0.1 M NaOH. The final solution was neutralized with HCl prior to analysis.

2.4 TOC analysis

All the TOC measurements of samples in this study were performed with the TOC-V CPH (Shimadzu Corporation) analyzer connected to an ASI-V autosampler (Shimadzu Corporation). The TOC analyzer was calibrated to the range of 0.5 to 20 mg/l. Quality control samples (QC WW4, 2.0 and 20.0 mg/l) were analysed at the beginning and at the end of each sample sequence. Prior to measurement, the samples were acidified with concentrated phosphoric acid.

2.5 Molecular size distribution

2.5.1 Analytical method

Size exclusion chromatography (SEC) is a widely used method for determining the molecular size distribution. It can provide simultaneously the number-average molecular weight (Mn) and the weight-average molecular weight (Mw), which are parameters used to describe the MWs of mixtures of molecules like HS. Since the molecular weights are strongly operational, the values calculated from the SEC measurements are "apparent molecular weights". In SEC the elution time depends of the molecular size and the elution of large molecules is faster than the elution of small molecules.

High performance SEC (HPSEC; Waters 2695 Separations Module) with ultraviolet (UV) detection (Hewlett Packard, Series 1050) at 254 nm (UV254) was used to investigate the molecular size distribution of isolated humic substances. Macroporous silica-based TSKgel G3000SW column (7.5 i.d. × 600mm) were used for the HS separation. The selected column has found to provide effective separation of the humic substances in HPSEC (Laborda et al. 2008, Peuravuori et al. 2005).

PSS standards with MWs 210, 1 100, 3 610 and 6 520 g/mol were used as calibration standards for the molecular size distribution investigations. PSS standards have been extensively employed in HS detection, even though they are less branched and cross-linked than humic substances (Asakawa et al. 2011, Asakawa et al. 2008, Guéguen and Cuss 2011, Laborda et al. 2008, Peuravuori and Pihlaja 1997). Nonetheless, the PSS standards have been considered to be suitable standards in studying the molecular weights of humic substances, since they behave similarly during chromatographic elution on HPSEC (Peuravuori and Pihlaja 1997, Chin et al. 1994). In our earlier studies changing the SEC conditions changed the elution time of HS and PSS rather equally (Luste et al. 2011).

Background, i.e. ionic strength, pH (≈7) and amount of phosphate buffer of standards, reference samples and groundwater samples were kept the same. The pH was not altered since the column suits best for the neutral pH or slightly acidic conditions (Janoš and

9

Zatřepálková 2007). The used eluent was 25% ACN in 10 mM phosphate buffer solution. Studies concerning the effect of the used eluent can be found in Luste et al. (2011). The flow rate of 0.6 ml/min was used and the sample volume was 40 µl. All the samples were filtered with 0.45 µm teflon filters prior to analysis.

2.5.2 Equations for calculatio

Molecular size distribution was determined with HPSEC. The calibration curve (log MW vs. peak retention time) was used to define the MW of the samples and Mn and Mw were further determined using Equations 1 and 2. In the equations, hi is the height of the measured peak eluted at retention time Ri and Mi is the molecular weight of the peak at retention time Ri, determined from the calibration curve. When the peak response exceeds 10 times the standard deviation of the average values of the baseline, it is considered as a peak (limit of quantification, LOQ). Peaks with significant intensity for HS were not found from the blank samples (MQ samples).

𝑀𝑀𝑛𝑛 = ∑ ℎ𝑖𝑖𝑛𝑛𝑖𝑖

∑ ℎ𝑖𝑖𝑀𝑀𝑖𝑖

𝑛𝑛𝑖𝑖

(1)

𝑀𝑀𝑤𝑤 = ∑ ℎ𝑖𝑖𝑀𝑀𝑖𝑖𝑛𝑛𝑖𝑖∑ ℎ𝑖𝑖𝑛𝑛𝑖𝑖

(2)

Polydispersity, ρ, measures the sample heterogeneity and can be calculated using Equation 3.

n

w

MM

=ρ (3)

2.6 ICP-MS/MS analysis method

ICP-MS/MS analysis was performed in order to examine the nature of metal complexation by humic substances and to detect a possible change in the concentration between different groundwater stations. The measurements were performed using 8800 Triple Quadrupole ICP-MS (Agilent). A set of elements were determined to find out the groundwater content and since some elements were thought to form complexes with HS or to be found in the HS structure based on literature. The measurements were done from both the original untreated samples and isolated (DAX-8) samples in ONK-PVA3, ONK-PVA5 and ONK-PVA10. In addition, a residue sample (water from DAX-8 isolation prosess that does not contain HS) was analysed.

The isolated samples were digested with HNO3 using Milestone Ethos One microwave since acidification of samples without digestion would lead to precipitation of HS in the sample. Blank samples were handled with the same procedure.

10

2.7 Results and discussion

2.7.1 Amount of HS

The results of the TOC measurements for the fraction of humic substances are presented at Table 1. The TOC content in isolated sample was expected to be HS. Previously the groundwater sample from the station ONK-PVA1 has been used as a sort of reference sample. However, sampling in ONK-PVA1 was not possible during the 2017 sampling and sample was taken from the station ONK-PVA3 instead. In this study humic substances covered around 30–40% on average of the TOC content in groundwater samples. TOC recoveries are similar to previous years’ studies (table 2). The results of isolated samples were calculated to equate the volume of the original sample. The results of QC samples prior and after each sample sequence at the TOC instrument, were always at the limits of laboratory’s accredited method.

Table 1. The results of the TOC measurements for the fraction of humic substances measured from HS isolation process and *Calculated for the original volume. TOC at the

original sample (mg/l)

TOC (HS) at isolated sample

(mg/l) *

TOC recovery (%)

ONK-PVA3 5.9 2.2 38

ONK-PVA5 2.2 0.72 33

ONK-PVA10 1.1 0.32 28

Reference 2.0 1.4 70

The amount of humic substances at ONKALO, Olkiluoto has been studied for many years. The current results have been compared with the previous ones at Table 2. Between the samplings, the amount of HS has remained at equal level in each groundwater station. According to the results, the amount of HS reduced when the sample was taken from the groundwater station situating further beneath the ground. This is reasonable since the amount of HS in the surface water is usually much greater compared to groundwater. Also, in the groundwater stations situating closer to the ground, humus originating from the surface water/soil is most probably leached into groundwater.

11

Table 2. The HS (mg/l) results at ONKALO, Olkiluoto in the previous and current study. Groundwater station Reference Sampling date Amount of HS (mg/l)

ONK-PVA1

Mäkelä et al. 2014 11.−14.11.2014 3.2

Luste et al. 2013 28.8.2012 3.4

Luste et al. 2011 14.10.2011 3.1

Vilhunen and Manninen 2010 3.6.2010 3.6

Manninen and Mäkelä 2006 n.a. 3.2

ONK-PVA3

Current study 23.3.2017 2.2

Luste et al. 2011 14.10.2011 2.3

Mäkelä and Manninen 2008 n.a. 2.8

Mäkelä and Manninen 2007 n.a. 3.0

ONK-PVA5

Current study 23.3.2017 0.72

Mäkelä et al. 2014 11.−14.11.2014 0.58

Luste et al. 2013 28.8.2012 0.65

Vilhunen and Manninen 2010 3.6.2010 0.75

ONK-PVA10

Current study 23.3.2017 0.32

Mäkelä et al. 2014 11.−14.11.2014 0.28

Luste et al. 2013 28.8.2012 0.35

n.a. = not available

2.7.2 Molecular size distribution The calibration was done using PSS standards 210, 1 100, 3 610 and 6 520 g/mol. The calibration curve is presented in the Figure 2.

Figure 2. The HPSEC calibration curve of PSS standards (MWs 210, 1 100, 3 610 and 6 520 g/mol).

12

The chromatograms of blank, ONK-PVA3, ONK-PVA5 and ONK-PVA10 samples are presented in Figure 3.

Figure 3. HPSEC chromatograms of blank, ONK-PVA3, ONK-PVA5 and ONK-PVA10.

The intensities of the sample from the station ONK-PVA10 were very low and the amount of humic substances seemed to be much lower in ONK-PVA10 sample compared to samples taken from ONK-PVA5 and especially from ONK-PVA3, according to the responses in chromatograms. However, the profiles (i.e. the shape of the curves with different intensities) in ONK-PVA5 and ONK-PVA10 were similar but the ONK-PVA3 profile had some differences when comparing to the other two samples. These differences refer the proportionally higher molecular weight compounds present in ONK-PVA3 sample compared to samples from ONK-PVA5 and ONK-PVA10. According to the chromatograms (Figure 3) the molecular size distribution appears to get lower as the sampling point goes further beneath the ground. Simultaneously the total amount of HS decreases.

The calculated Mn and Mw values for each groundwater sample are presented in Table 3 (calculations using Equations 1 and 2). The values were slightly lower when the samples were from the groundwater station situating further beneath the ground. However, no differences were found between the ONK-PVA5 and ONK-PVA10 values. When exploring the results, it should be noted that the HS in the groundwater sample does not represent the original situation but the isolated HS in completely different environment. Results of the reference samples (Luste et al. 2013) have shown that the treatments have some effect on the results by slightly reducing the values. Thus, also in case of the groundwater samples the expected values are somewhat higher than the measured ones.

13

Table 3. Molecular size distribution results for ONK-PVA3 and ONK-PVA5 and results from the studies in 2014 and 2013.

ONK-PVA1 ONK-PVA3 ONK-PVA5 ONK-PVA10

Mn Mw Mn Mw Mn Mw Mn Mw Current study 335 804 321 791 320 794

Mäkelä et al 2014 418 879 386 557 529 547

Luste et al 2013 790 1 020 523 633 443 537

The values for polydispersivity, i.e. sample heterogeneity, are presented in Table 4 (calculated using Equation 3). According to the results, polydispersivity was found to be somewhat higher in the current study compared to the previous years’ results. In addition, no difference was found between the ONK-PVA3, ONK-PVA5 and ONK-PVA10 samples. Polydispersivity is a measure of variability in molecular size distribution of a sample. According to three year studies, significant differences in polydispersivity values were not observed (measurement uncertainty taken account).

Table 4. Polydispersivity (ρ) values for ONK-PVA3 and ONK-PVA5 samples and results from previous studies in 2014 and 2013.

ONK-PVA1 ONK-PVA3 ONK-PVA5 ONK-PVA10 Current study 2.4 2.5 2.5 Mäkelä et al 2014 2.1 1.9 1.1 Luste et al 2013 1.3 1.2 1.2

2.7.3 Analysis of metal content in groundwater samples by ICP-MS/MS Elemental composition of HS typically consists of carbon, hydrogen, oxygen, nitrogen, sulfur and trace elements (Frimmel 2005). Hetero elements like oxygen, nitrogen and sulfur are found in several functional groups such as carboxyl groups, phenolic and carbohydrate hydroxyl groups, amino groups, quinonic groups and reactive positions in aromatic structures (Frimmel 2005, Klavins et al. 2012). Due to the different functionalities HS have an ability to interact with metal ions and have an impact on their mobility, behavior and speciation forms in the environment.

After the isolation of humic substances with DAX-8 metal concentrations were measured from the samples taken from three different groundwater stations ONK-PVA3, ONK-PVA5 and ONK-PVA10 using Agilent 8800 ICP-MS triple quad. The isolated samples were digested with HNO3 using Milestone Ethos One microwave since acidification of samples without digestion would lead to precipitation of HS in the sample. For comparison, metals were also measured from the original untreated samples. Since the metal binding capacity of humic substances is strongly pH-dependent (Frimmel 2005) the metal concentrations in the isolated groundwater samples may not be transferred directly to the original samples. Therefore the metal recovery in HS is an estimate since the metal binding capacity of HS may have changed during the isolation procedure.

Aquatic HS is known to contain sulfur ranging from 0.5 to 1.43% (Xia et al. 1998) in the structure and therefore sulfur concentration could be used as a tool to estimate the change in the amount of HS between different groundwater stations.

14

Table 5 present the results for elemental determinations of both the original (i.e. the untreated samples) and isolated HS samples. In addition, the residue samples, i.e. water that came through the DAX-8 resin without HS, was analysed for comparison. Values detected in blank samples were decreased from the results. The elements which are present in the original sample in large extent are also seen in humic fraction as the highest amounts. However, the correlation is not clear and presence of some elements as impurity can not be ruled out. Table 5. Elemental analyses from original and HS samples. Li B Mg P S K Ca Mn Fe Cu Sr Ba

ONK- PVA3

Original (µg/l) 20 220 18 000 72 79 000 7 400 54 000 170 62 0.51 520 2.4

Residue (µg/l) 22 280 21 000 80 85 000 8 300 57 000 180 98 1.75 530 2.5

HS fract. (µg/l)* 0.17 7.2 78 0.57 780 66 340 0.59 1.3 0.024 3.6 0.012

Recovery (%) 0.86 3.3 0.43 0.79 0.99 0.90 0.64 0.35 2.1 4.7 0.69 0.50

ONK- PVA5

Original (µg/l) 19 780 13 000 <2.0 39 000 3 200 250 000 76 <10 <0.50 2 600 12

Residue (µg/l) 20 880 14 000 2.2 42 000 3 400 250 000 79 21 1.45 2 600 11

HS fract. (µg/l)* 0.14 11 75 0.06 280 24 1 500 0.35 <10 <0.50 16 0.11

Recovery (%) 0.76 1.5 0.58 n.a. 0.71 0.76 0.61 0.46 n.a. n.a. 0.61 0.90

ONK-PVA10

Original (µg/l) 24 1 300 20 000 <2.0 150 9 300 510 000 120 <10 1.3 4 500 42

Residue (µg/l) 24 1 400 21 000 3.3 120 9 800 520 000 120 24 1.2 4 700 41

HS fract. (µg/l)* 0.18 14 15 <2.0 7.8 69 2 400 0.06 <10 <0.50 28 0.34

Recovery (%) 0.75 1.1 0.08 n.a. 5.2 0.75 0.48 0.05 n.a. n.a. 0.62 0.81

*Calculated for initial volume Sulfur concentration of the isolated samples (HS) decreased by 99% when comparing ONK-PVA3 and ONK-PVA10 samples and sulphur concentration in ONK-PVA5 was in between these to values. This supports the results obtained from TOC and SEC analyses and indicate that the HS content decreases when going further beneath the ground and that the HS content in ONK-PVA10 is extreamly low. However, major and parallel differences has been in sulfur concentration in initial sample as well. It is to be noted that some changes in the sulfur content might also be due to changes in HS structure between different groundwater stations and extraction procedure. Some elemental concetrations decreased and some increased when going further beneath the ground. With some elements, the concentration was the lowest in sampling point ONK-PVA5. The recoveries have been mainly the same between the samples from different groundwater stations. In very low concentrations (near the limit of quantification) the relative inaccuracy is higher.

15

2.8 Conclusions

The annual follow-up of HS in the ONKALO groundwater was continued. As a result of the isolation method the average amount of humic substances were 2.2 mg/l in ONK-PVA3, 0.72 mg/l in ONK-PVA5 and 0.32 mg/l in ONK-PVA10. These results are similar to the previous studies. The amount of HS covered around 30–40% of the TOC content in groundwater samples. The amount of HS in the groundwater appears to be lower at the water from the groundwater stations situating further beneath the ground.

The molecular size distribution measurements in order to find out the MW, Mn and Mw of the humic matter were performed using HPSEC with UV254 detection. The eluent was 25% ACN in 10 mM phosphate buffer. The apparent Mn values of ONK-PVA3, ONK-PVA5 and ONK-PVA10 water were around 335, 321 and 320, respectivley, whereas the Mw values were 804, 791 and 794. No significant variations according to sampling depth were found in SEC studies in 2017. However, the sampling at ONK-PVA1 was not possible in 2017 and the results have differed the most in ONK-PVA1 compared to other groundwater stations. Altogether, according to the TOC measurements the amount HS was considerably lower in ONK-PVA5 and ONK-PVA10 samples compared to the ONK-PVA3 sample.

Elemental concentrations were also measured from the three groundwater samples in order to evaluate the change in water chemistry between different groundwater stations. The decrease in sulfur concentration in isolated groundwater samples (HS fraction) when going further beneath the ground was found to support the results of TOC analysis.

16

17

3. COLLOIDS

3.1 Sampling preparations and sampling procedures The collection of colloid samples was carried out during 12.–13.12.2017. The samples were taken from the groundwater stations ONK-PVA5, ONK-PVA10 and ONK-PVA11. Figure 1 illustrates the location of the groundwater stations in ONKALO.

The colloid samples were collected on a Nuclepore polycarbonate 0.05 µm filter. Three parallel samples were taken simultaneously and the volume of water that ran through the filters was measured.

The sample lines were (polytetrafluoroethylene) PTFE ¼ in. (inner diameter) lines and the valves as well as the fittings were Swagelock’s (PTFE). The sample containers were made of plastic (high density polyethylene, HDPE). The filter holders were made of polycarbonate.

The filter holders were cleaned by pre-treatment with ultra clean Millipore water with an ultrasonic mixer. The filters and holders were assembled at the Eurofins Environment Testing laboratory and sealed before the sampling to avoid any contamination by aerosol particles.

Once on site, the filtering lines were flushed with the sample water before sampling.

Filtration of ONK-PVA5 samples took less than 2 hours (~1 000 ml) and in ONK-PVA10 around 3 hours (~600 ml). The on-line pressure in ONK-PVA11 was lower and the sampling took over 21 h time. During that time the bottles were overflown and the volume of the water ran through the filters (~1 700 ml) was estimated noticing the volume of the overflown but collected water. After the sampling the filter holders were vacuum-dried and sealed. The sealed filter holders were sent to Mikrofokus Oy for SEM analysis to investigate the composition of the colloid material on the filters.

3.2 Single particle analysis method SEM micrographs were taken from the filters at Mikrofokus Oy in Helsinki. Major enhancement in SEM method compared to previous years was higher amount of micrographs per filter. The number of the micrographs was 24 per sample, i.e. 72 per groundwater station. Also the magnification factor 2000x was used. The micrographs were taken randomly from the filter surface with magnification factors of 1 000x, 2 000x, 4 000x and 8 000x. In addition, the used equipmet was replaiced by one with much higher micrograph resolution. The image sizes were 2 560 * 1 920 pixels and resolution was 508 dpi

The micrographs were delivered to Eurofins Environment Testing as jpg-image files. According to the SEM micrographs and the analyst, colloid particles were fairly equally distributed. The particle calculation from the micrograph images was done with computer software ImageJ. In the previous studies (Järvinen et al., 2011 and Luste et al., 2014) software proved to be useful for the calculation of colloidal particles from SEM micrograps.

The particle calculation process initiated with a step of turning the SEM micrograph images into binary images. The binary images were needed for the software to be able to

18

detect the particles. The calculation boundaries were set so that particles smaller than 5x5 pixel were not calculated. The upper boundary was set to 100x100 pixels. The circularity factor was set to 0.01–1, where 1 is a perfectly round particle and when the value gets closer to 0 it indicates increasingly elongated shape.

The circularity factor is calculated as following (equation 4):

4𝜋𝜋 × 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑃𝑃𝐴𝐴𝐴𝐴𝑃𝑃𝑃𝑃𝐴𝐴𝑃𝑃𝐴𝐴𝐴𝐴2

(4)

ImageJ calculates the particle areas and the output gives results in pixel values. Based on the area measurements, the diameter of the particles was then calculated by equation 5. Particles, with diameter less than 2 µm were explored.

𝑑𝑑 = 2𝑥𝑥�𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝜋𝜋

(5)

Diameters in pixels were then transformed into metric values and the conversion factors for each magnification are presented in table 6.

Table 6. Scaling pixels to µm in different magnifications. Magnification Scale: 1 pixel to µm

1 000 1 pixel = 0.05 µm 2 000 1 pixel = 0.025 µm 4 000 1 pixel = 0.0125 µm 8 000 1 pixel = 0.00625 µm

From this data, the particle size distribution (number of particles in size group per liter) for size classes within the range of 0.05–2.0 µm was produced. Particle and mass concentrations of colloids were estimated similarly to the previous study by Luste et al. (2014). The particle concentration (pt/ml) for each size interval was calculated by the following equation 6:

VsddN

dcolld

⋅⋅

=∑

∑i

-1

sØ

Ø][ (6)

where Σ corresponds to the sum of the micrographs examined, dN/dØ is the number of detected particles in size group dØ (µm), s (µm2) is the active surface area of the filter, si (µm2) is the surface area of the micrograph, and V is the filtered volume (l).

The mass concentration (mg/l) for different size intervals was determined by the following equation 7:

Ø][

6Ø

Ø][ 3

dcolld

dcolld

⋅⋅⋅

=πρ (7)

where ρ is the particle density (mg/mm3) and Ø is the particle diameter (µm). The density of the particles was assumed to be 2 (mg/mm3) and the shape was assumed to be spherical.

19

The particles on the micrographs are rarely spherical, and so the mass concentration is overestimated. A correction factor was not used.

3.3 Results

The colloid particles were fairly equally distributed in the SEM micrographs on ONK-PVA5 and ONK-PVA10 filters and SEM micrographs of different magnifications also appeared similar in these filters. According to SEM micrographs, in ONK-PVA11 filters material was layered on the filters and they were almost totally blocked. Since the particles were in layers, calculation of results was impossible.

Only few random bacteria were observed from the micrographs. Figures 4–6 present SEM micrographs of different groundwater stations.

Figure 4. SEM micrographs of ONK-PVA5 (magnifications of 1 000x and 4 000x).

20

Figure 5. SEM micrographs of ONK-PVA10 (magnifications of 1 000x and 8 000x).

Figure 6. SEM micrograph of ONK-PVA11 (magnifications 8 000x).

Results on total particle concentration as well as particle and mass concentration for < 1 µm size class of colloids in groundwater during 2013–2017 from ONK-PVA1, ONK-PVA3, ONK-PVA5, ONK-PVA10 and ONK-PVA11 is shown in Table 7. In 2017 ONK-PVA5 and ONK-PVA10 particle concentrations were higher but in comparable level to the previous study. The increase in concentration can origin from the actual sample or might be due to the enhanced SEM apparatus providing remarkably higher resolution and onwards higher amount of noticed particles. Since the layering effect in ONK-PVA11 filters, results of the year 2017 can not be provided.

21

Table 7. Total particle concentration and particle and mass concentration for < 1µm size class of colloids in groundwater extracted from ONK-PVA1, ONK-PVA3, ONK-PVA5, ONK-PVA10 and ONK-PVA11.

Total particle concentration (particles/l)

Particle concentration (pt/l) (< 1µm size class)

Mass concentration (mg/l)(< 1µm size

class) PVA1 (2013) 1.0E+09 8.6E+08 0.163 PVA3 (2014) 1.3E+08 1.1E+08 0.024 PVA5 (2013) 1.1E+08 1.1E+08 0.010 PVA5 (2017*) 9.0E+08 8.9E+08 0.046

PVA10 (2013) 1.3E+08 1.2E+08 0.014

PVA10 (2017*) 7.4E+08 7.3E+08 0.030

PVA11 (2015) 2.1E+08 1.8E+08 0.042

*Current study Overall, the colloid particle distribution in ONK-PVA5 and ONK-PVA10 were somewhat different to previous years’ results (Figure 7). In 2017 majority of colloids ranged within size range of 0.05−0.5 µm. In previous study at 2013 small particle were detected in much lesser extent, which might be due to less precice micrographs. The most dominant particle size in ONK-PVA5 was 0.2 µm in 2013 and 2017. In ONK-PVA10 the most dominant particle size in 2013 was 0.1 µm but in 2017 0.3 µm.

Dirrefences between the groundwater sampling stations were thus found. Particles were smaller in groundwater station situating further beneath the earth surface.

22

Figure 7. Colloid particle concentration in groundwater from ONK-PVA5 and ONK-PVA10 in 2013 and 2017. Colloid mass concentration differed also from the previous study at 2013 (Figure 8). As mass concentration, the particle size around 0.4 µm dominated in both groundwater stations ONK-PVA5 and ONK-PVA10. Previously in 2013, larger particles around 0.8–1 µm have been dominant. This difference might also origin from the enhancements in SEM method producing micropraphs of higher resolution. The mass concentration in ONK-PVA5 was higher compared to ONK-PVA10 sample.

23

Figure 8. Colloid mass concentration in groundwater from ONK-PVA5 and ONK-PVA10 in 2013 and 2017.

24

3.4 Conclusions

In this study the concentration of groundwater colloids at the groundwater station ONK-PVA5, ONK-PVA10 and ONK-PVA11 was evaluated and the results were compared with the previous ones from 2013–2015. Sampling in ONK-PVA11 however failed, and results could not be provided. Overall concentrations in ONK-PVA5 and ONK-PVA10 samples were present in similar extent as in earlier study.

Overall, the colloid particle distribution in ONK-PVA5 and ONK-PVA10 were somewhat different to previous results. In 2017, smaller particles within size range of 0.05−0.5 µm were observed compared to the similar study in 2013. The most dominant particle size in ONK-PVA5 was 0.2 µm and in ONK-PVA10 0.1 µm. Particles were found to be smaller in ONK-PVA10 compared to ONK-PVA5, i.e. smaller particles are present in larger extent whith groundwater stations situating further beneath the earth surface.

Colloid mass concentration differed also from the previous study at 2013. As mass concentration, the particle size around 0.4 µm dominated in both groundwater sampling stations ONK-PVA5 and ONK-PVA10, which differes from the previous study (2013: around 0.8–1 µm). The mass concentration in ONK-PVA5 was higher compared to ONK-PVA10 sample. Differences compared to previous years are expected to origin from the enhancements in SEM method producing micropraphs of higher resolution.

25

RECOMMENDATIONS

The study of both HS and colloids has been stabilized during the years and the next follow-up study for humic substances is suggested to be performed in 2018 and for colloids in 2019. The proposed study set-up is the same or very similar to the year 2017.

26

27

REFERENCES

G. R. Aiken, Isolation and concentration techniques for aquatic humic substances, in: G. R. Aiken, D. M. McKnight, R. L. Wershaw, P. MacCarthy (Eds.), Humic substances in soil, sediment and water: Geochemistry, Isolation and Characterisation, Wiley, NY, 1985, pp. 363-385. D. Asakawa, Y. Imura, T. Kiyota, Y. Yanagi and N. Fujitake; 1218, 2011, 6448-6453 "Molecular size fractionation of soil humic acids using preparative high performance size-exclusion chromatography" D. Asakawa, T. Kiyota, Y. Yanagi and N. Fujitake; Analytical Sciences, 24, 2008, 607-613 "Optimisation of conditions for high-performance size-exclusion chromatography of different soil humic acids" Y.-P. Chin, G. Aiken and E. O’Loughlin; Environmental Science and Technology, 28, 1994, 1853-1858 “Molecular weight, polydispersity, and spectroscopic properties of aquatic humic substances” C. Guéguen and C. Cuss; Journal of Chromatography A, 1218, 2011, 4188-4198 "Characterisation of aquatic dissolved organic matter by asymmetrical flow field-flow fractionation coupled to UV-Visible diode array and excitation emission matrix fluorescence" F. Frimmel, 2005, Biopolymers Online - Wiley Online Library “Aquatic Humic Substances” M. Klavins , O. Purmalis, V. Rodinov, 2013, Estonian Journal of Ecology, “Peat humic acid properties and factors influencing their variability in a temperate bog ecosystem” P. Janoš and I. Zatřepálková; Journal of Chromatography A, 1160, 2007, 160-165 "High-performance size-exclusion chromatography of humic substances on the hydroxyethyl methacrylate column" E. Järvinen, P. Manninen, M. Takala, and S. Vilhunen, (2011); Sampling and Characterisation of Groundwater Colloids in ONKALO at Olkiluoto, Finland, 2009-2010. Posiva Working Report 2011-29. F. Laborda, E. Bolea, M. Górriz, M. Martín-Ruiz, S. Ruiz-Beguería and J. Castillo; Analytica Chimica Acta, 606, 2008, 1-8 "A speciation methodology to study the contributions of humic-like and fulvic-like acids to the mobilisation of metals from compost using size exclusion chromatography-ultraviolet absorption-inductively coupled plasma mass spectrometry and deconvolution analysis" J. A. Leenheer; Environmental Science & Technology, 15, 1981, 578-587 “Comprehensive approach to preparative isolation and fractionation of dissolved organic carbon from natural waters and wastewaters”

28

S. Luste, J. Mäkelä, P. Manninen; Posiva Working Report 2011 "Comparable investigation of the molecular size distribution and the amount of humic substances isolated from ONKALO, Olkiluoto, 2011"

S. Luste, J. Mäkelä, P. Manninen; Posiva Working Report 2013 "Comparable investigation of the molecular size distribution and the amount of humic substances isolated from ONKALO, Olkiluoto, 2012"

S. Luste, M. Takala, P. Manninen; Posiva Working Report 2014-28 “Sampling and characterisation of groundwater colloids in ONKALO at Olkiluoto, Finland, 2013” R. L. Malcolm; Analytica Chimica Acta, 232, 1990, 19-30 “The uniqueness of humic substances in each of soil, stream and marine environments” P. Manninen and J. Mäkelä; Posiva Working Report 2006-42 “Humic and fulvic acids in groundwater – A preliminary research” J. Mäkelä and P. Manninen; Posiva Working Report 2007-23 “Humic and fulvic acids in groundwater” J. Mäkelä and P. Manninen; Posiva Working Report 2008-36 “Molecular size distribution and structure investigations of humic substances in groundwater” J. Mäkelä, M. Mäkelä, K. Koikkalainen, P. Manninen; Posiva working report 2016 “Study of humic substances and dissolved organic matter isolated from the groundwater of, Olkiluoto, 2014” J. Peuravuori, A. Monteiro, L. Eglite and K. Pihlaja; Talanta, 65, 2005, 408-422 ”Comparative study for separation of aquatic humic-type organic constituents by DAX-8, PVP and DEAE sorbing solids and tangential ultrafiltration: elemental composition, size-exclusion chromatography, UV-vis and FT-IR” J. Peuravuori and K. Pihlaja; Analytica Chimica Acta, 337, 1997, 133-149 “Molecular size distribution and spectroscopic properties of aquatic humic substances” J. Peuravuori, K. Pihlaja and N. Välimäki; Environment International, 23, 1997, 453-464 “Isolation and characterisation of natural organic matter from lake water: Two different adsorption chromatographic methods” E. M. Thurman and R. L. Malcolm; Environmental Science & Technology, 15, 1981, 463-466 “Preparative isolation of aquatic humic substances” R. Vidali, E. Remoundaki and M. Tsezos; Aquatic Geochemistry, 16, 2010, 207-218 "Humic acid copper binding following their photochemical alteration by simulated solar light"

29

S. Vilhunen and P. Manninen; Posiva Working Report 2010-80 “Comparable investigation of the molecular size distribution and the amount of humic substances isolated from ONKALO, Olkiluoto 2010” K. Xia, F. Weesner, W. Bleam, P. R. Bloom, U. Skyllberg, and P. Helmke., 1998, Soil Science Society of America Journal Abstract “XANES Studies of Oxidation States of Sulfur in Aquatic and Soil Humic Substances”

30

Wo

rk

ing

Re

po

rt 2

01

4-4

4 •

Re

su

lts o

f Mo

nito

ring

at O

lkilu

oto

in 2

01

3 - H

yd

rog

eo

ch

em

istry