MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Miljø- og Fødevareudvalget 2021-22
MOF Alm.del Bilag 717
Offentligt

HITLIST4

Non-targeted and

suspect screening of

sewage sludge

[Serietype og nummer]

August 2022

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Publisher: The Danish Environmental Protection Agency

Editors: Martin Hansen, Mulatu Y. Nanusha, Emil Egede Frøkjær, Martin Mørk Larsen &

Jens Søndergaard

This scientific report has been reviewed by four Steering Group Members from the Danish

Environmental Protection Agency: Maj-Britt Bjergager, Ida Rasmussen, Jakob Bruun Nico-

laisen and Helle Rüsz Hansen.

Graphics: Department of Environmental Science, Aarhus University

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Contents

Abstract

Acronyms

2.1

2.2

2.3

2.4

3.1

3.2

3.3

3.4

3.5

Introduction

Methodology

Sewage sludge samples

Analytical platforms

Sample preparation

Data analysis

Results and discussion

Inorganic elements

NTS dataset

Identified organic substances

Suspects

Multivariate data analysis and data exploration

Conclusion

References

Sample overview

Methods

Internal standards

ICP-MS dataset

NTS dataset

Appendix 1.

Appendix 2.

Appendix 3.

Appendix 4.

Appendix 5.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Abstract

This project applied non-targeted and suspect screening to activated sludge from five Danish

wastewater facilities. In addition, the concentrations of 61 elements were also determined in

the sludge.

The sludge concentrations showed relatively high levels of mercury (0.4-1.4 mg/kg) and

cadmium (0.4-1.4 mg/kg), compared to natural levels in arable soils. Also, a high copper (203-

571 mg/kg) and zinc (502-1616 mg/kg) concentrations in sludge relative to arable soil show

that the use of sludge is likely to increase the level of copper and zinc over time in sludge

amended fields.

Five non-targeted screening platforms were applied to analyse the sewage sludge. After

data analysis, thousands of substances were discovered in the non-targeted dataset. Search-

ing the data against in-house and international databases 41 substances were determined at

the highest annotation level (1), 1,751 molecules were identified at the second highest level

(2), 7,091compounds at level 3, and a total of 15,471 combined at level four and five. Exam-

ples of confirmed substances are 1H-benzotriazole, 2,6-dichlorophenol, bisphenol S,

methylparaben, terbutryn and prosulfocarb.

By applying novel semi-quantitative concepts, it was possible to predict concentrations of

513 substances, e.g. PFOS was detected at all five sites with concentrations ranging from 2.4-

47.4 µg/kg sludge, while 6PPD-quinone was only found in three sites (14-74 µg/kg). Endoge-

nous or natural metabolites were found, e.g. bile acids as deoxycholate (58-3247 µg/kg) and

peptides as Gly-Leu-Lys (53-938 µg/kg).

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Acronyms

ESI

HLB

HRMS

ICP-MS

LOD

MFS

nLC

NTS

PFAS

PFCs

ppb

ppm

SPE

Dry weight

Electron impact ionization (GC)

(Heated) electrospray ionisation (LC)

Gas chromatography

Hydrophilic-lipophilic balanced polymer

High-resolution mass spectrometry

Inductively coupled plasma mass spectrometry

Liquid chromatography

Limit of detection

Miljøfarlige stoffer (substances of emerging concern)

Mass spectrometry

nano-liquid chromatography

Non-targeted screening analysis

Per- and polyfluoroalkyl substance

Perfluorochemicals

Parts-per-billion (ng/mL, µg/L)

Parts-per-million (µg/mL, mg/L)

Quality control

Solid phase extraction

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

1. Introduction

Sewage sludge is rich in nutrients and can be applied to agricultural soils (BEK nr 1001 af

27/06/2018). However, sludge may also contain elevated levels of contaminants such as heavy

metals that can have adverse effects on the environment and pose a health risk for human

consumption of the agricultural produce. Organic micropollutants, or substances of environmen-

tal concern (termed MFS in Danish) is another category of contaminants that may enter the

environment via agricultural sludge amendment to exert unwanted public and environmental

health side-effects.

Recently, it was demonstrated that, holistic non-targeted screening analysis (NTS) provides

the means to perform mass suspect screening and go beyond to discover previously unknown

molecular entities in environmental samples. NTS is revolutionary and fundamentally different

from targeted monitoring strategies and has a large potential for effective evaluation of water

quality regulations. NTS is based on high-resolution mass spectrometry (HRMS), that rapidly

profile thousands of (unknown) substances in complex environmental samples [1]–[3]. The NTS

strategy is used when former unknown compounds are detected in a sample and data is inves-

tigated without any presumptions or knowledge of the sample [4]. Suspect screening is another

strategy used for searching HRMS data for known chemicals, i.e. by using a reference list of

pesticides and biocides which are expected to be present in the sample [4]. As such, NTS is

used to describe this entire field of research.

The NTS concept was recently developed and applied in a research project (HITLIST1) un-

der the Danish Environmental Protection Agency’s Pesticide Research Program [1]. The work

was followed by a recent Danish EPA research project (HITLIST2) that mapped the detectable

chemical space and further validated the NTS methodology, so it can be used as a reliable

monitoring method for the NOVANA water quality program. Finally, a third research report is

currently under review (HITLIST3) on applying the NTS concept on Danish surface water sam-

ples.

The aim of the present project ‘HITLIST4’ was to apply NTS and ICP-MS methodologies to

determine inorganic and organic micropollutants in sewage sludge. A secondary objective was

to compile a list of substances identified at each location, make concentration estimations based

on novel semi-quantitative approaches and use multivariate statistics for in-depth data explora-

tion.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

2. Methodology

2.1

Sewage sludge samples

A total of five facilities provided activated sludge samples in November and December 2021

(Table 1). In all cases sludge were stored in anaerobic digesters following by dewatering. All

sludges are approved for use on agricultural soils (grade A).

Table 1.

Overview of studied wastewater sludge samples and facility details. PE, actual per-

son equivalents burden.

Facility

Type

Digestion

Roskilde

Måløv

Ejby Mølle

Egå

Herning

Biodenipho

Biodenipho (modified)

Biodenipho

125.000

70.000

225.000

90.000

150.000

3 weeks anaerobic digestion

3-4 weeks anaerobic digestion

4 weeks anaerobic digestion

3 weeks anaerobic digestion

Thermophilic digestion (2 weeks) fol-

lowed by mesophilic digestion (2

weeks)

2.2

Analytical platforms

Five high-resolution mass spectrometry NTS analytical platforms were used to analyse the sam-

ples for organic micropollutants. In addition, an accredited inductively coupled plasma mass

spectrometry (ICP-MS) methodology was used to determine concentrations of 61 inorganic el-

ements in the sludge. The five NTS platforms were reverse-phase nano-liquid chromatography

electrospray ionisation high-resolution tandem mass spectrometry (nLC-ESI-HRMS/MS in pos-

itive and negative ionisation modes), cap-flow cLC-ESI-HRMS/MS methods (in positive and neg-

ative ionization mode) directed towards detecting e.g. perfluorochemicals and gas chromatog-

raphy electron impact ionisation high-resolution mass spectrometry (GC-EI-HRMS) and are fur-

ther detailed in Appendix 2.

nLC-ESI(+)-HRMS/MS

nLC-ESI(-)-HRMS/MS

cLC-ESI(+)-HRMS/MS

cLC-ESI(-)-HRMS/MS

GC-EI(+)-HRMS

ICP-MS

2.3

Sample preparation

Samples were collected in Rilsan bags and immediately stored in cooler box (5

C) and trans-

ported to the analytical lab and stored at -20

C. Sample aliquots of 0.20 grams for inorganic

element analysis were acid extracted using microwave extraction and analysed by ICP-MS ac-

cording to EPA method 3051A. Sample aliquots of 0.20 grams for organic micropollutant analy-

sis were spiked with isotope-enriched internal standards and extracted using pressurized liquid

extraction in a two-way workflow (one for LC and another for GC). Sample extracts for GC-

HRMS workflows were analysed directly, while extracts for LC-HRMS workflows were further

purified using solid-phase extraction (for details see Appendix 2).

2.4

Data analysis

The LC and GC-HRMS raw data were processed using Compound Discoverer 3.3 (Thermo

Fisher) for peak detection, retention time alignment and peak picking for non-target screening.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

An overview of the Compound Discoverer and non-target data processing workflow can be

found in (Appendix 2.6). The output of this is a feature list, i.e. a table with m/z and retention

time pairs (features) and their peak area, which were further processed for the identification and

structural elucidation of contaminants. The detected peaks were prioritized based on the criteria

such as peak intensity threshold, blank subtraction, reasonable peak symmetry (sharp peak

apex), molecular formula predicted from the exact mass and the isotopic pattern as well as

structural similarity match with the analytical reference standard. The data analysis filtration and

decision tree(s) are available in detail in Appendix 2.7. The features/compounds were confirmed

to different identification levels (level 1 to 5) as suggested by Schymanski et al. [5]. In short, a

level 1 substance has been confirmed by a reference standard on the same NTS platform (MS,

MS/MS and retention time matching). A level 2 substance is a highly probable structure matched

to literature and/or public MS/MS spectral libraries, and may further be supported by retention

time predictions. A level 3 substance is tentative candidate structure where no MS/MS libraries

exist, however the experimental evidence is supported by in silico MS/MS fragmentation predic-

tions. A level 4 substances is when an unequivocal chemical formula can be assigned from MS

spectral data (e.g. adduct, isotopic pattern). A level 5 substance is when only the exact mass

(m/z) can be assigned. Concentrations of level 1 and 2 identified substances were performed

by novel semi-quantitative approaches [6], and are described in detail in Appendix 2.8. Statistical

analyses were performed using GraphPad Prism 9.3 and MetaboAnalyst 5.0.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

3. Results and discussion

3.1

Inorganic elements

The ICP-MS analyses included screening of 61 elements of the five sludge samples (Figure

1). Specific element concentrations across the five sites were relatively similar, e.g. arsenic

(As) varied from 2.99 in Roskilde to 8.26 mg/kg dw in Ejby Mølle and mercury (Hg) from 0.41

in Ejby Mølle to 1.41 mg/kg dw in Måløv. Iron (Fe) was observed in the highest concentration

(46.2 g/kg dw in Måløv), while rhenium (Re) was found at the lowest levels (0.002 mg/kg dw in

Ejby Mølle and Måløv). A complete and detailed dataset of all 61 elements is available in Ap-

pendix 4.

Figure 1.

Concentrations of 61 inorganic elements in wastewater activated sludge (mg/kg dry

weight) determined with ICP-MS.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

A comparison with limit values in BEK nr 1001 of 27/06/2018 for inorganic elements in waste

applied for agricultural or private gardens use are compiled in Table 2. As shown, all five

sludge samples comply with the safety limit values. The limit values expressed on a dry weight

basis were exceeded for cadmium and/or mercury (and just exceeded for nickel) in three of

the samples but were below the limit value expressed per amount phosphorus. According to

BEK nr. 1001, the waste must comply with the limit values expressed on either dry weight ba-

sis or phosphorus basis, so all samples comply with this criterion. However, the results indi-

cate that the metals of particular concern in the waste are cadmium and mercury, which

should be monitored closely especially if large amounts of waste are spread or if a significant

variation in composition is expected.

Comparison with average levels from 430 samples of arable soil [7] in 1992 and 1993 indicate

levels in the sludge is within a factor of two for arsenic, between 1.4 and 7 times higher for

cadmium, chromium, nickel and lead, whereas mercury is a factor of 8-28 times higher in

sludge than in arable soil. Copper and zinc are used in many household products and leaks

from plumbing, so levels are 17-73 times higher in sludge than arable soils. It has been ar-

gued, that the sludge amendment leads to increasing levels of copper and zinc in arable land

topsoil (0-25 cm) of respectively 0.16 and 0.96 mg/kg/y, or 1.7 and 3% increase per year from

1998 to 2014, and slightly less below the ploughing zone (25-50 cm) at 1.6 and 2.8% increase

per year [8].

Compared to the levels in the Earth’s Crust particularly bismuth and gold are concentrated

in the sludge samples (270-610 resp. 45-110 times above Earth Crust levels), but also anti-

mony, selenium, phosphorous, silver, zinc, mercury and platinum are present in up to 10-40

times the level in the earth crust (in declining order). Bismuth present at very low levels in the

Earth Crust (around 0.009 mg/kg), but is used both for medicine, in personal care products, as

a catalyst for rubber and fibers, in alloys and have been used in shot and bullets as replace-

ment for lead. The use of gold in electronics, medical treatments and nano-gold are potential

sources for the sludge. Antimony is also used in electronics, as alloys (mainly with lead), and

antimony compounds are used for flame-retardant materials, paints, glass and pottery. Sele-

nium is used as additive to glass and in photocells, solar cells and photocopiers, it is toxic and

used in anti-dandruff shampoos, but it is also an essential element needed in small amounts,

with both carcinogenic and teratogenic effects at raised levels. Silver is used in jewelry and

utensils, in mirrors and solder, electrical contacts and batteries. It was previously used in pho-

tography, and now as antibacterial nano-particles in e.g. clothes, foot ware and touchscreen-

enabled gloves. Platinum is also used for jewelry, but today mainly as catalytic converters for

cars, trucks and busses. Also, other industrial catalytically uses, in electronics including optical

fibers and LCDs, and in chemotherapy drugs. All uses are cited from the Royal Society of

Chemistry [9]. The high ratio metals can mostly be attributed to industrial, therapeutic or

household product use of rare metals or metalloids, whereas the more ordinary elements like

phosphor and zinc are used in large quantities by households, farmers and industry to give the

overrepresentation in sludge. Mercury is supposed to be more or less banned, and the high

levels in sludge are mainly attributed to legacy usage. In Europe it is probably mainly used in

fluorescence lighting, the majority is from long range transport from artisan goldmining in Af-

rica, together with pollution from coal burning in powerplants in Eastern Europe and Asia.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Table 2.

Limit values of inorganic elements in waste for agricultural use and private gardens in

BEK no. 1001 of 27/06/2018 in comparison to measured values in the five sludge samples.

The sludge must comply with either the limit value for an element in mg kg

-1

dry weight or the

limit value in mg kg

-1

total phosphorus (P). Limit values for inorganic elements in soil in BEK no

1001 and the average concentrations in Danish soils are listed in the bottom of the table for

comparison. A detailed table is available in Appendix 4.

Concentrations in mg kg

-1

dry weight

Safety limit values for

waste used for agricul-

ture (BEK no. 1001,

27/06/2018)

Safety limit values for

waste used for private

gardens (BEK no.

1001, 27/06/2018)

N001 Roskilde (n=2)

N004 Ejby Mølle

N007 Egå

N010 Måløv

N013 Herning (n=2)

Safety limit values for

soil (BEK no. 1001,

27/06/2018)

Average soil level

1992/3 N=430 (Larsen

et al, 1996)

Concentrations in mg kg

-1

total P

100

1000

4000

0.8

120

2500

100

200

10000

100

1000

335

203

348

571

214

4000

926

502

693

736

1616

0.8

1.4

0.4

0.8

0.7

1.1

0.8

1.0

0.4

0.6

1.4

0.6

2500

605

363

410

556

1061

100

200

5000

1235

640

491

550

1228

100

0.5

0.2

0.05

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

3.2

NTS dataset

After data analysis pipelines were applied (Appendix 2.6) to the NTS HRMS dataset (LC and

GC), nearly half a million features were observed, and they could be grouped into thousands of

substances. After lab procedural blank background filtration, the GC EI dataset was assembled

into

2,300 substances whereas 1,056 substances were annotated at level 2, 189 compounds

at level 3 and 1027 compounds at level 5. The combined LC datasets contained over 20,000

features across both positive and negative ionization modes. Nearly 15,000 of these could only

be annotated as either m/z or as a calculated chemical formula (levels 5 and 4 respectively).

6,902 compounds (not taking duplicates into account) were annotated at level 3 across both

platforms, i.e. with proposed chemical structure based on MS2 fragments. A total of 725 unique

compounds were identified at level 2 across both platforms, with 546 detected in positive mode,

215 in negative mode, and 36 compounds detected in both platforms. 29 compounds were de-

tected at level 1 using positive ionization and 15 using negative, with a combined total of 41

unique level 1 compounds across the two platforms (LC negative and LC positive).

The current state-of-the-art semi-quantitative methods can predict concentrations of LC (ESI)

data, however not GC (EI) data and more research efforts are urgently needed in this area.

3.3

Identified organic substances

The identity of 41 substances were confirmed (level 1) across the dataset (Table 2). See Ap-

pendix 2.8 for a detailed description of data curation for level 1 annotation. The complete dataset

with level 1-5 annotations is available through Appendix 5.

Table 3.

Substances identified at the highest level (1) with observed minimum and maximum

sludge concentrations (across all identified platforms) and detection frequencies (D

). n.c. de-

noted concentration not calculated by the semi-quantitation algorithm.

Name

min

(µg/kg)

Industrial substances

1H-Benzotriazole

4-Methyl-1H-benzotriazole

5-Methyl-1H-benzotriazole

Bisphenol S

Dimethyl phthalate

Ethylparaben

Methylparaben

Tributyl phosphate

Triisobutyl phosphate

Tris(2-butoxyethyl) phosphate

Vanillin

204

189

287

n.c.

127

639

Pesticides and biocides

2,6-Dichlorophenol

Clomazone

Prosulfocarb

Terbutryn

n.c.

Pharmaceuticals

Aspartame

Azithromycin

Caffeine

Carbamazepine

103

n.c.

685

2807

9075

n.c.

100

n.c.

305

654

3283

617

1577

116

964

952

343

n.c.

750

17431

100

max

(µg/kg)

(%)

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Cetirizine

Citalopram

Cotinine

Fexofenadine

Furosemide

Lamotrigine

Losartan

Metoprolol

Miconazole

Nicotine

Propranolol

Salicylic acid

Sertraline

Venlafaxine

Diclofenac

Ibuprofen

114

n.c.

104

Natural products

1120

n.c.

189

n.c.

496

1020

n.c.

809

449

4048

362

599

100

Daidzein

Genistein

PFAS substances

2962

Perfluorodecanoic acid (PFDA)

Perfluorooctanesulfonic acid (PFOS)

Perfluorononanoic acid (PFNA)

Perfluorooctanesulfonamide (PFOSA)

100

3.4

Suspects

A list of suspects was used for screening throughout the dataset. The combined dataset also

revealed the tyre residues 6PPD and 6PPD-quinone were present in sludge samples. The two

substances were detected using two platforms, i.e. demonstrating the importance of combining

NTS platforms to obtain a more complete chemical fingerprint.

It is expected the applied NTS platforms will be able to detect all listed suspects of interest

(Table 3), however substances could be present below detection limits or hampered by matrix

interference. Some substances are identified to level 2 as no in-house reference standard was

available. Other substances as level 3 (tentative candidates), as the obtained MS/MS-frag-

mentation spectral score was low.

Table 4.

Suspects extracted from the NTS dataset.

Compound name

Industrial substances

6PPD

6PPD-Quinone

1,2,4-Triazole

Pesticides and biocides

Trimethoprim

Pyrethrin

Cypermethrin

Yes

Found

Annotation level

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Permethrin

Deltamethrin

Lamda-cyhalothrin

Spinosad A

Pharmaceuticals

beta-Estradiol

(+)-Estrone

Azithromycin

Citalopram

Clarithromycin

Diclofenac

(-)-Erythromycin

Naproxen

Ibuprofen

2-hydroxyibuprofen

Propranolol

Tramadol

Carbamazepine

Telmisartan

Metformin

Metoprolol

Sertraline

Ciprofloxacin

Venlafaxine

PFAS substances

Trifluoroacetic acid

Heptafluorobutyric acid (PFBA)

Perfluoropentanoic acid (PFPeA)

Perfluorohexanoic acid (PFHxA)

Perfluoroheptanoic acid (PFHpA)

Perfluorooctanoic acid (PFOA)

Perfluorononanoic acid (PFNA)

Perfluorodecanoic acid (PFDA)

Perfluorododecanoic acid (PFDoA)

Perfluorotridecanoic acid (PFTrDA)

Perfluorobutanesulfonic acid (PFBS)

Perfluorohexanesulfonic acid (PFHxS)

Perfluoroheptanesulfonic acid (PFHpS)

Perfluorooctanesulfonic acid (PFOS)

Perfluorodecanesulfonic acid (PFDS)

Perfluorooctanesulfonamide (PFOSA)

6:2 FTSA (1h,1h,2h,2h-perfluorooctanesulfonic acid)

Yes

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

3.5

Multivariate data analysis and data exploration

The analysed NTS data revealed very different molecular fingerprints between wastewater treat-

ment plants (Figure 2). Clearly, the sewage sludge samples fall into three groups having very

similar molecular signatures; Zealand (Roskilde, N01-02 and Måløv, N10-12); Fuenen (Ejby

Mølle, N04-06) and Jutland (Egå, N07-09 and Herning, N13-15).

Figure 2:

Scores (red) and loadings (blue) in biplot using multivariate principal component anal-

ysis is used for data exploration of the samples (red) and concentrations of 513 substances

(blue). Samples clustered in proximity to each other has very similar chemical profiles, whereas

samples far from each other are very dissimilar. Substances in proximity with specific samples

are highly associated with its occurrence at this site. For simplicity substances has been num-

bered from 1-513 and table with substance names are found in Appendix 5.

Evidently, some chemicals are highly associated with sewage sludge from specific sites. E.g.

the industrial chemical 1-Methyl-1H-benzotriazole (substance no. 74 in proximity to sample

N10) is in high abundance in Roskilde and Måløv sludges (305-504 µg/kg), yet still occurring in

the Herning sludge, albeit at lower concentrations (140-181 µg/kg, Appendix 5). Some chemi-

cals are common across the samples and located at the centre region of the PCA biplot, e.g.

Mono(2-ethylhexyl) phthalate (MEHP, substance no. 402 and is present in all samples at simi-

lar levels; 20-57 µg/kg, Appendix 5).

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Figure 3:

Scores and loadings plot in multivariate principal component analysis is used for data

exploration of the samples analysed in GC EI dataset. Each sample is displayed in the scores

plot (left). Samples clustered in proximity to each other has very similar chemical profiles,

whereas samples far from each other are very dissimilar. The loadings plot (right) displays every

detected substance present in the given dataset (1,056 substances annotated at level 2 for GC

IE). Overlaying the two plots the association between substance and sample is visualised. One

sample (Roskilde) were omitted as outlier.

Figure 4:

Differential analysis between Herning and Ejby Mølle for GC-EI dataset. Each dot is

a substance and analysis are performed with n=3 for each group. Substances on left side (blue

area) is in higher abundance in Herning when compared to Ejby Mølle. Substances on right side

(red area) is in lower abundance in Herning when compared to Ejby Mølle. The higher (vertical

y-axis) adjusted -log p value to more significant. 238 substances (blue) are at significantly higher

abundance in Herning and another 227 substances (ren) are significantly higher abundance in

Ejby Mølle, while 168 substances (grey) are not significantly different in abundance between the

two sites. One of the encircled substances (red ring) is sertraline (GC-2103, see Appendix 5 for

substance names).

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

4. Conclusion

In terms of inorganic elements, all sludge samples complied with the criterion in the current BEK

no. 1001 on waste applied for agriculture. The metals causing most concern in sludge are rela-

tively high levels of mercury and cadmium, compared to natural levels in arable soils. But also

interesting to see the high levels of particular rare metals as bismuth and antimony, and precious

metals like gold, silver and platinum, all of which with many uses in both industry, as catalysts

and alloys, as therapeutic products in hospitals and in house hold products like jewellery, elec-

tronics, nano-particles in clothes and cosmetic products. The high copper and zinc concentra-

tions in sludge relative to arable soil show that the use of sludge is likely to increase the level of

copper and zinc over time in sludge amended fields, both in the top soil and below the ploughing

zone.

The NTS dataset revealed that a great number of micropollutants are present in waster sludge

and the chemical fingerprint varies across wastewater sites. Combined with suspect screening

and semi-quantitative concentration determinations it was possible to estimate sludge concen-

trations of more than 500 chemicals. To name a few substances observed at all sites, diclo-

fenac with concentrations ranging 4-70 µg/kg, ethylparaben (287-953 µg/kg), perfluorodeca-

noic acid (PFDA, 2-11 µg/kg), azithromycin (11-2807 µg/kg) and tributyl phosphate (93-343

µg/kg). The chemical identity and occurrence were very different between sites, exemplified by

the herbicide prosulfocarb observed at three out five sites at 61-305 µg/kg. Several parent

chemicals and associated metabolites were also observed, such as venlafaxine and O-

desmethyl-venlafaxine. The dataset also revealed five different per- and polyfluoroalkyl sub-

stances are present in sewage sludge. Especially, PFOS and PFDA were omnipresent and

detected in all samples at 2-53 µg/kg.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

5. References

[1]

[2]

“Danish EPA pesticide and biocide research programme, project: HITLIST, grant no.

MST-667-00207, project period: June 2018 to December 2019.”

T. K. O. Gravert, J. Vuaille, J. Magid, and M. Hansen, “Non-target analysis of organic

waste amended agricultural soils: Characterisation of added organic pollution.,” Under

review, 2021.

N. L. Ma

et al.,

“Body mass, mercury exposure, biochemistry and untargeted

metabolomics of incubating common eiders (Somateria mollissima) in three Baltic

colonies,”

Environ. Int.,

vol. 142, no. January, p. 105866, 2020.

J. R. Sobus

et al.,

“Integrating tools for non-targeted analysis research and chemical

safety evaluations at the US EPA,”

J. Expo. Sci. Environ. Epidemiol.,

vol. 28, no. 5, pp.

411–426, 2018.

E. L. Schymanski

et al.,

“Identifying small molecules via high resolution mass

spectrometry: Communicating confidence,”

Environmental Science and Technology,

vol. 48, no. 4. pp. 2097–2098, 2014.

J. Liigand, T. Wang, J. Kellogg, J. Smedsgaard, N. Cech, and A. Kruve,

“Quantification for non-targeted LC/MS screening without standard substances,”

Sci.

Rep.,

vol. 10, no. 1, p. 5808, Dec. 2020.

M. M. Larsen, J. Bak, and J. Scott-Fordsmand, “Monitering af tungmetaller i danske

dyrknings- og naturjorde : Prøvetagning i 1992/93.”

J. Jensen, M. M. Larsen, and J. Bak, “National monitoring study in Denmark finds

increased and critical levels of copper and zinc in arable soils fertilized with pig slurry

*,”

Environ. Pollut.,

vol. 214, pp. 334–340, 2016.

“Royal Society of Chemistry.” .

J. P. Koelmel

et al.,

“Acquisition With Automated Exclusion List Generation,”

J Am Soc

Mass Spectrom.,

vol. 28, no. 5, pp. 908–917, 2018.

A. Schlosser and R. Volkmer-Engert, “Volatile polydimethylcyclosiloxanes in the

ambient laboratory air identified as source of extreme background signals in

nanoelectrospray mass spectrometry,”

J. Mass Spectrom.,

vol. 38, no. 5, pp. 523–525,

2003.

J. V. Olsen

et al.,

“Parts per million mass accuracy on an orbitrap mass spectrometer

via lock mass injection into a C-trap,”

Mol. Cell. Proteomics,

2005.

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Appendix 1.

Sample overview

Table 5.

Overview of sludge samples and locations. The legends/names used in data analysis

are underlined.

Site

FORS Roskilde Spildevand A/S

Vandcenter Syd, Ejby Mølle

Århus vand, Egå renseanlæg

Novafos Måløv Renseanlæg

Herning Vand

Sample ID (n)

N001, N002, N003

N004, N005, N006

N007, N008, N009

N010, N011, N012

N013, N014, N015

Sampling date

November 24, 2021

December 14, 2021

December 13, 2021

December 14, 2021

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Appendix 2.

Appendix 2.1

Methods

Liquid chromatography high-resolution mass spectrometry

This system was used for two NTS platforms (nLC-ESI(+)-HRMS/MS and nLC-ESI(-)-

HRMS/MS). Nano-liquid chromatographic separation was performed on a Dionex Ultimate 3000

NCS-3500RS Nano Proflow system (Thermo Scientific). Ready samples were stored in glass

96-well plates in a Dionex WPS-3000 TPL RS autosampler at 8°C. Sample were loaded (1 µL)

onto a nanoflow UHPLC column (PepMap RSLC, C18, 2 µm, 100 Å, 50 µm x 150 mm, Thermo

Scientific) equipped with a titanium inline filter frit (0.5 µm). The flow rate of mobile phases was

300 nL/min. Chromatographic separation was achieved using a gradient beginning at 10 % mo-

bile phase B (98 % acetonitrile, 2 % water, and 0.1 % formic acid) and 90 % mobile phase A (2

% acetonitrile, 98 % water, and 0.1 % formic acid) kept for 2 minutes. Thereafter the gradient

increased to 95 % over 15 minutes. This level of 95 % B was kept for another 10 minutes. The

conditions were restored to 10 % mobile phase B over 0.5 minutes followed by 12.5 minutes of

equilibration time, leading to a total runtime of 40 minutes. In between each injection the needle

and fluidics were washed with 200 µL of 80 % acetonitrile and 0.1 % formic acid in water. The

pump systems were rinsed every hour with a seal wash solution of 10 % methanol and 0.1 %

formic acid in water. All solvents used were of UHPLC-MS grade. The mass spectrometric anal-

ysis was performed on a high-resolution tandem mass spectrometer (Q Exactive HF, Thermo

Scientific). Analytes were ionised by electrospray ionisation using an EASY-Spray ion source.

The applied spray voltage was 1.50 kV during positive polarity and1.70 kV during negative po-

larity with a capillary temperature of 250 °C and an S-lens RF level of 50. No sheath, aux, and

sweep gas was used.

HRMS acquisition was done in either full scan mode for quantification or iterative data-depend-

ent fragmentation (ddMS

) mode for identification [10]. Both the positive and negative polarity

modes were used. Full scan acquisition was recorded using a resolution of 240K at m/z 200, an

automatic gain control (AGC) target of 1e6, a maximum injection time of 100 ms, and a scan

range of 70-1050 m/z for positive mode and 100-1500 m/z for negative mode. ddMS

acquisition

was done using full scan settings with a resolution of 240K, AGC target of 1e6, maximum IT of

100 s, and scan range of 120-1500 m/z at m/z 200 for positive mode, and 100-1500 m/z at m/z

200 for negative mode. ddMS

settings used a resolution of 15K, maximum IT of 50 s, an isola-

tion window of 1.0 m/z, AGC target of 5e4, loop count of 5, and stepped collision energies of 30,

70, and 120 NCE. The acquisition was performed with a dynamic exclusion of 5 s, minimum

AGC target of 500, charge exclusion of >2, and an apex trigger between 2-6 s. An estimated

chromatic peak width (FWHM) was set to 3 s. Sub-ppm mass accuracy was ensured by real

time calibration of a lock mass of 371.10124 (polysiloxane from air) during positive polarity and

112.98563 (sodium formate cluster) during negative polarisation [11], [12]. Calibration of the

mass spectrometer was performed with Pierce™ LTQ Velos ESI Positive and Negative Ion Cal-

ibration Solutions (Thermo-Fisher Scientific).

Instrumental performance was ensured by regular monitoring of an in-house laboratory quality

control sample prepared from fetal bovine serum.

Appendix 2.2

Gas chromatography high-resolution mass spectrometry

GC-HRMS analysis was achieved using an Orbitrap mass spectrometer (Exactive GC, Ther-

moFisher Scientific) with a TriPlus autosampler and a TraceGOLD TG-5MS analytical column

(30 m, 0.25 µm, 0.25 mm, 5% phenyl - 95% dimethyl polysiloxane phase, ThermoFisher Scien-

tific) installed in a TRACE 1310 GC (ThermoFisher Scientific). As described in the HITLIST2

report, one-microliter sample extract was injected sandwiched with air using a split-splitless

mode at 280 °C and 70 mL/min split flow after 60 sec. The column was operated with high purity

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

helium at 1.00 mL/min and a temperature program; initial 60 °C with 2 min hold and ramped (5

°C/min) to 240 °C and further (10 °C/min) to 300 °C with a final holding time of 16 min. Analytes

were transferred using a MS-transferline at 280 °C and ionized using electron impact ionisation

(EI) at 70 eV with a 12 minutes filament delay. The Orbitrap HRMS system was operated in full

scan mode (m/z 50 to 750) at a 60,000 resolution in centroid mode and an automatic gain-

control target of 1e6 ions. The Q Exactive HRMS system was tuned and calibrated on a daily

basis using FC43.

Appendix 2.3

CapLC-HRMS/MS for PFC analysis

Same system as described under Appendix 2.1 was utilized, however retrofitted with micro-bore

UHPLC analytical column for PFC analysis. Chromatographic separation was performed on a

Dionex Ultimate 3000 NCS-3500RS high-flow system (Thermo Scientific). Ready samples were

stored in plastic vials in a Dionex WPS-3000 TPL RS autosampler at 8°C. Sample were loaded

(1 µL) onto an UHPLC column (Phenomenex, C18, 3 µm, 100 Å, 75 µm x 150 mm) equipped

with a 20 mm guard column with same material and inner diameter. The flow rate of mobile

phases was 1000 nL/min. Chromatographic separation was achieved using a gradient beginning

at 10 % mobile phase B (10 mM NH

Ac in 60 % MeCN) and 90 % mobile phase A (10 mM

Ac in 10% MeCN) kept for 2 minutes. In between each injection the needle and fluidics were

washed with 200 µL of 80 % acetonitrile and 0.1 % formic acid in water. The pump systems

were rinsed every hour with a seal wash solution of 80 % methanol and 0.1 % formic acid in

water. All solvents used were of UHPLC-MS grade. The mass spectrometric analysis was per-

formed as described in Appendix 2.1.

Appendix 2.4

Sample preparation for inorganic element analysis (ICP-MS)

Sample aliquots of 0.2 grams for inorganic element analysis were acid extracted using inverse

aqua regia (6 ml Merck Suprapure HNO

+ 2 ml Merck Suprapure HCl) in an Anton Paar Multi-

wave 7000 microwave oven according to EPA method 3051A and subsequently diluted with MQ

water and analysed by an Agilent 7900 ICP-MS for 61 elements. For quality assessment and

control, 3 procedural blanks, 2 duplicate samples (for Roskilde and Herning sludge) and 6 Cer-

tified Reference Material samples (3 ERM-CC144 and 3 IMEP-21) were digested and measured

with the samples. Elemental analysis under the NOVAVA programme is done by ICP-MS and

includes the whole periodic table from lithium (third element) to uranium (92

element). Ele-

ments above uranium; neptunium, plutonium and americium are radioactive but the first two can

be found in low concentrations in uranium ores, and the remaining elements from 95 to 118

have only been synthesised in labs or during nuclear fission testing. Some elements is not ana-

lysed by ICP-MS due to use as internal standards (typically rhodium, iridium and indium), or

used as plasma source and collision cell (argon, helium) or as digestion acids (hydrogen, oxy-

gen, nitrogen, chloride and for total dissolution fluoride and boron). The elements normally mon-

itored in NOVANA are mercury, cadmium, lead, nickel, chromium, arsenic, copper, silver (in

biota) and zinc.

Appendix 2.5

Sample preparation for organic micropollutant analysis

Sample aliquots of 0.2 grams were mixed with 5 grams of pre-washed diatomaceous earth and

placed in 10 mL pressurized liquid-extraction cells pre-fitting with glass fibre filters. Cells were

spiked with internal standards. Capped PLE cells were extracted twice in two extracts; two cy-

cles with methanol:water (1:1) followed by two cycles with dichloromethane:hexane (1:1). Pre-

heat time was 5 minutes, purge volume 60% and purge time 60 seconds. Methanol:water ex-

tracts were directed towards LC-HRMS analysis, while DCM:hexane extracts were prepared for

GC-HRMS analysis. Methanol:water extracts were prior to LC-analysis purified using 500 mg

HLB solid-phase material.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Appendix 2.6

Quality control samples

Four types of quality control samples were prepared; 1) laboratory procedural blanks by per-

forming entire sample preparation workflow without adding any sample matrix. These samples

would be used for background filtration. 2) A pooled sample by combining all sample extracts.

3) A composite sample were made by aliquoting 0.5 grams of each sample into a pool. 4) Two

certified reference materials of ICP-MS analyses; ERM-CC144 and IMEP-21. Calibration stand-

ards were prepared by adding 0, 50, 100, 150, and 200 ng internal standards (see Table 7) into

five aliquots of 200 µL total pooled quality control sample respectively. A calibration blank was

prepared by adding 50 ng of internal standards into 200 µL methanol.

Appendix 2.7

Post-processing

After acquiring NTS raw data for non-target screening, the bottleneck was the identification or

assigning the correct chemical structure for the features. Here, the feature (the combination of

m/z and retention time) represents a particular compound in the sample. Compound Discoverer

version 3.3, a commercial software package developed by Thermo Fisher Scientific, was used

for peak detection, retention time alignment and peak picking. The workflow displaying the se-

lected processing nodes and the associated workflow connections is given in Figure 4. The

general overview of the NTS workflow used in

Compound Discoverer

for the raw data pro-

cessing.. The raw files obtained in full scan mode (samples, blanks and pooled QCs) and data-

dependent, MS/MS, mode (pooled QCs) were processed. In the workflow, the pooled QCs were

labelled as “identification only” which were used as a source of fragmentation data. The main

preliminary data processing workflow nodes includes input files, select spectra, align retention

times, detect compound and mark backgrounds nodes. The list of features for the ionized com-

pounds present in the samples, blanks and pooled QCs were created by the “Detect Com-

pounds” node. Then, the generated ion list was used by the “Group Compounds” node which

combines chromatographic peaks across the raw files by using their molecular weight and re-

tention time. Afterwards, the “Predict compositions” node predicts elemental compositions for

all features/compounds, which are subsequently annotated against compounds whose chemical

information was previously recorded in mzCloud, ChemSpider (exact mass or formula) and local

database searches against Mass Lists (exact mass with or without RT). “Assign compound an-

notation” node performs spectral similarity search against mzCloud (online database, ddMS2

and/or DIA) and mzVault (inhouse spectral database), for compounds with ddMS2. Finally, the

“Mark Background Compounds” node incorporates blanks to trace features/compounds arising

from sample preparation.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Figure 5.

The general overview of the NTS workflow used in

Compound Discoverer

for the

raw data processing.

The output of this was a feature list, i.e. a table with m/z and retention time pairs (features) and

their peak area as well as other necessary parameters depending on the nodes used. The fea-

ture was subject to peak prioritization for the purpose of identification and structural elucidation

of contaminants. Peak intensity threshold, blank subtraction, reasonable peak symmetry (sharp

peak apex), reasonable elemental composition predicted from the exact mass and the isotopic

pattern as well as structural similarity match were among the feature prioritization criteria used

(nontarget data processing workflow, Figure 4). Every step of the nontarget workflow leads from

lower identification confidence (level 5) to higher (level 1) suggested by Schymanski et al., 2014.

Briefly, the identification journey was started with the HRMS features with exact masses (level

5), whose unequivocal chemical formulas were computed based on the isotopic pattern of peaks

and adducts (level 4). The plausibility of computed formulas was evaluated by searching against

the online chemical data (e.g. ChemSpider) and in-house-mass list. To move on to level 3, the

tentative chemical structure was searched against the online spectral library through compound

discoverer (e.g. mzCloud) and manual search against MetFrag, SIRUIS and MassBank), and

in-house library (mzVault) using triggered MS2 fragmentation data. Here, tentative candidates

that match MS1 accurate mass and the MS2 fragmentation spectra were identified. The diag-

nostic MS/MS fragment masses and/or ionization behaviour together with the information on

parent compounds were used to categorize the tentative candidates to the plausible/probable

chemical structure (level 2). Then, the identity of the compound was confirmed by comparing it

with MS/MS fragmentation spectra of the analytical reference standard (level 1).

GC EI HRMS were search against NIST and NORMAN libraries. Substances with total spec-

trum scores above 70 were assigned as level 2 identifications, while scores below this value are

assigned as level 3. Substances without chemicals formulas were assigned as level 5.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Figure 6.

Nontarget data processing workflow for the categorization and confirmation of fea-

tures to identification confidence levels.

Appendix 2.8

Quality assurance of level 1 data

Before compounds were confidently annotation at level 1, these entries were manually curated

according to the workflow below:

A signal-to-noise ratio greater than 5 between at least 1 sample and a corresponding

procedural blank.

Matching retention time (Δt

< 0.1 min) between the compounds identified in a quality

control sample and spiked quality control sample as well as a larger peak area in the

spiked quality control sample compared to the unspiked quality control sample.

Matching MS

data between sample and the spectral reference entry.

No ambiguity between the compound of interest and isobaric compounds. In the case

where a compound peak cannot be precisely defined on the basis of both retention

time and MS

spectrum due to isobaric interferences, its annotation must be down-

graded to at least level 2.

Appendix 2.9

Semi-quantitative predictions

The concentration of identified compounds was estimated by employing the Semi-Quantifica-

tion tool. The ionization efficiency prediction approach that accounts for structural similarity

with standard compounds was used. The CSV file containing SMILES, retention time, signal

(peak area), and standard compounds with known concentrations were subject to the semi-

quantification software. Here, a representative sample (a quality control sample) with a high

number of detects was selected for the concentration estimation. Candidate compounds not

detected in the selected quality control sample were not quantified. Thereby, a total of 513

compounds were subjected to quantification. The concentration for unknown compounds in

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

the samples was calculated based on the estimated concentration for the quality control sam-

ple using equation (1).

��

equation (1)

��

Where concentration estimated by Semi-Quant, max peak area used for concentration estima-

tion, concentration of the compound in the individual sample and peak area of the compound in

the sample.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Appendix 3.

Internal

standards

For use in semi-quantification and as quality assurance of the implemented acquisition work-

flows, several isotope labelled extraction and internal standards were used in the study. An ab-

solute amount of 50 ng of each isotope labelled extraction standards (see Table 4) were added

to each of the 15 PLE packed soil samples before extraction. Only 10 ng of GC-EI standards

were added. After PLE, an additional 50 ng of internal standards (Table 5) were added to the

extracts of all samples. A pooled composite sample was used as post, pre, and non ES-spiked

sample respectively to calculate total (PLE+SPE) extraction recoveries.

Table 6.

Recoveries of isotope labelled extraction standards. N.D., not detected.

Platform

GC-EI

Standard

1,3,6,8-Tetrachloro(13C12)dibenzo-p-dioxin

2,3,7,8-Tetrachloro(13C12)dibenzo-p-dioxin

1,2,3,7,8-Pentachloro(13C12)dibenzo-p-dioxin

1,2,3,4,7,8-Hexachloro(13C12)dibenzo-p-dioxin

1,2,3,6,7,8-Hexachloro(13C12)dibenzo-p-dioxin

1,2,3,7,8,9-Hexachloro(13C12)dibenzo-p-dioxin

1,2,3,4,6,7,8-Heptachloro(13C12)dibenzo-p-dioxin

*Octachloro(13C12)dibenzo-p-dioxin

2,3,7,8-Tetrachloro(13C12)dibenzofuran

1,2,3,7,8-Pentachloro(13C12)dibenzofuran

2,3,4,7,8-Pentachloro(13C12)dibenzofuran

1,2,3,4,7,8-Hexachloro(13C12)dibenzofuran

1,2,3,6,7,8-Hexachloro(13C12)dibenzofuran

1,2,3,7,8,9-Hexachloro(13C12)dibenzofuran

2,3,4,6,7,8-Hexachloro(13C12)dibenzofuran

1,2,3,4,6,7,8-Heptachloro(13C12)dibenzofuran

1,2,3,4,7,8,9-Heptachloro(13C12)dibenzofuran

Octachloro(13C12)dibenzofuran

3,3',4,4'-Tetrachloro(13C12)biphenyl

3,4,4',5-Tetrachloro(13C12)biphenyl

2,3,3',4,4'-Pentachloro(13C12)biphenyl

2,3,4,4',5-Pentachloro(13C12)biphenyl

2',3,4,4',5-Pentachloro(13C12)biphenyl

2,3',4,4',5-Pentachloro(13C12)biphenyl

2,3,3',4,4',5-Hexachloro(13C12)biphenyl

2,3,3',4,4',5'-Hexachloro(13C12)biphenyl

2,3',4,4',5,5'-Hexachloro(13C12)biphenyl

3,3',4,4',5,5'-Hexachloro(13C12)biphenyl

2,2',3,3',4,4',5-Heptachloro(13C12)biphenyl

Retention

time (min)

40.69

42.35

45.44

48.14

48.24

48.46

50.91

N.D.

41.90

44.54

45.17

47.45

47.56

48.00

48.72

49.98

51.3

53.51

39.19

39.57

40.33

40.48

40.83

41.37

42.62

43.24

44.02

44.21

45.41

44.62

43%

58%

57%

58%

31%

56%

33%

25%

40%

10%

42%

44%

49%

47%

45%

42%

48%

44%

47%

49%

54%

41%

Recovery

(%)

50%

49%

11%

19%

22%

<1%

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

2,2',3,4,4',5,5'-Heptachloro(13C12)biphenyl

2,3,3',4,4',5,5'-Heptachloro(13C12)biphenyl

nLC-pos

13C3-Caffeine

13C6-Thiabendazole

DEET-D7

Diuron-D6

Imidacloprid-D4

Pirimicarb-D6

nLC-neg

2,4-D-D3

Dicamba-D3

Diuron-D6

Imidacloprid-D4

Mecoprop-D3

cLC-neg

Perfluoro-n-(2,3,4-[13]C3)butanoic acid (

-PFBA)

Perfluoro-n-(1,2-[13]C2)octanoic acid (

-PFOA)

Perfluoro-n-1-(1,2,3,4-[13]C4)octanesulfonate (

PFOS)

Perfluoro-n-(1,2-[13]C2)decanoic acid (

-PFDA)

45.6

46.65

14.47

13.81

22.80

23.13

17.87

16.07

22.76

21.10

23.20

17.84

24.08

4.93

5.70

6.44

6.45

36%

42%

88%

69%

68%

82%

87%

76%

94%

<1%

105%

88%

109%

<1%

76%

148%

174%

Table 7.

Signal stability (%RSD) of isotope labelled internal standards. The instrumental %RSD

is calculated from peak areas of repeated injections (n=12) of quality control samples, measured

every 4th injection throughout the instrument acquisition.

Platform

nLC-pos

Standard

13C3-Testosterone-2,3,4

13C4-15N2-Riboflavin

Retention time (min)

23.78

14.89

%RSD

2.67

5.60

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Appendix 3.1

Signal stability of extraction and internal standards

The signal stability was calculated from the total average of normalised values of all standards

measured in each sample:

��

− ��

∑(

)

��

max ��

��=1

��

Where

is the peak area of standard

in sample

i, µ

is the mean area of standard

across all

samples respectively,

max

is the maximum peak area for standard

across all samples, and N

is the number of standards identified.

100

-20

-40

-60

-80

-100

ESTD signal deviation from QC mean (%)

Procedural blank

Ejby_1

Ejby_2

Ejby_3

Egå_1

Egå_2

Egå_3

Måløv_1

Måløv_2

Måløv_3

QC_01

QC_02

QC_03

QC_04

QC_05

QC_06

QC_07

QC_08

QC_09

QC_10

QC_11

Herning_1

Herning_2

Roskilde_1

Roskilde_2

Herning_3

Post_spike

Pre_spike

Figure 9.

Average signal difference of the six isotope labelled extraction standards in each

sample from the mean of twelve QC injections on the LC-pos system (13C3-caffeine, DEET-

D7, 13C6-thiabendazole, diuron-D6, pirimicarb-D6, and imidacloprid-D4). Error bars cor-

respond to 1��½�.

The Danish Environmental Protection Agency / HITLIST2

Non-spike

QC_12

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

100

-20

-40

-60

-80

-100

ISTD signal deviation from QC mean (%)

Procedural blank

Ejby_1

Ejby_2

Ejby_3

Egå_1

Egå_2

Egå_3

Måløv_1

Måløv_2

Måløv_3

QC_01

QC_02

QC_03

QC_04

QC_05

QC_06

QC_07

QC_08

QC_09

QC_10

QC_09

QC_11

QC_10

Herning_1

Herning_2

Roskilde_1

Roskilde_2

Herning_3

Post_spike

Pre_spike

Figure 10.

Average signal difference of the two isotope labelled internal standards in each

sample from the mean of twelve QC injections on the LC-pos system (13C3-Testosterone-

2,3,4 and 13C4-15N2-Riboflavin(-)). Error bars correspond to 1��½�.

100

ESTD signal deviation from QC mean (%)

-20

-40

-60

-80

-100

Procedural blank

Egå_1

Egå_2

Ejby_1

Ejby_2

Roskilde_1

Roskilde_2

Ejby_3

Egå_3

Måløv_1

Måløv_2

Måløv_3

Herning_1

Herning_2

Herning_3

Non-spike

QC_01

QC_02

QC_03

QC_04

QC_05

QC_06

QC_07

QC_08

QC_11

QC_12

Pre_spike

Post_spike

Figure 11.

Average signal difference of the four isotope labelled extraction standards in each

sample from the mean of twelve QC injections on the LC-neg system (Diuron-D6, imidacloprid-

D4, mecoprop-D3, and 2,4-D-D3). Error bars correspond to 1��½�.

Non-spike

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Appendix 4.

ICP-MS dataset

A complete ICP-MS dataset for 61 elements is available via …MST.dk.

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

Appendix 5.

NTS dataset

The complete NTS dataset is available via …MST.dk.

File “Appendix 5-HITLIST4-GCEI.xlsx” contains all substances from the GC-EI-HRMS dataset.

File “Appendix 5-HITLIST4-LC.xlsx” contains all substances from the LC-HRMS/MS datasets

(nLC and cLC).

The Danish Environmental Protection Agency / HITLIST2

MOF, Alm.del - 2021-22 - Bilag 717: Orientering om resultaterne af projekt om undersøgelse af miljøfarlige forurenende stoffer i slam

HITLIST4: Non-targeted and suspect screening of sewage sludge

The Danish Environmental

Protection Agency

Tolderlundsvej 5

DK - 5000 Odense C