Per- and polyfluoroalkyl substances (PFAS) profiles in primary and secondary landfill leachates: Indications of transformation, liner interactions, and other PFAS sources

Abstract

Per- and polyfluoroalkyl substances (PFAS) are a growing concern due to their persistence, bioaccumulation potential, and continued widespread use in consumer products. PFAS disposed of in landfills are emitted to the environment via leachate, which drives a need to better understand PFAS behavior in landfills and landfill liner systems. This study examines PFAS concentrations in primary and secondary leachate from three municipal solid waste landfills utilizing double HDPE geomembrane liner systems. Samples were also analyzed for physical-chemical constituents such as chloride, ammonia, chemical oxygen demand, and metals. On average, physical-chemical parameter concentrations were significantly lower in the secondary compared to the primary leachate, although PFAS concentrations were not significantly different between leachate sources. Concentrations of chloride in groundwater and primary leachate were used to calculate expected PFAS concentrations in the secondary leachate. PFAS concentrations in secondary leachate were often higher than expected, with PFAAs more likely to exceed expected levels. Of the 92 PFAS analyzed, 50 were quantified in primary leachates and 48 in secondary leachates. The ΣPFAS concentrations in primary leachate ranged from 3200–81,000 ngL⁻¹, and secondary leachate ranged from 3300–96,000 ngL⁻¹. Possible explanations for the disproportionately high PFAS concentrations in secondary leachates, including residence time, transformation, liner sorption, and other PFAS sources (e.g., landfill gas) are explored. While liner systems are highly effective, PFAS migration through landfill liners and potential groundwater impacts remain a concern. This study underscores the importance of continued research into PFAS migration mechanisms and the potential environmental impacts of unidentified precursor PFAS in landfills.

Graphical Abstract:

1. Introduction

Per- and polyfluoroalkyl substances (PFAS) are a class of persistent organic chemicals employed in numerous consumer and industrial products for their surfactant, stick-resistant, stain-resistant, water-resistant, and grease-resistant properties [10,18,20]. The extensive use of PFAS in consumer products, combined with the significant PFAS concentrations leaching from products during use and end-of-life management [2,6,7,24,26,31,30,47], and the documented health effects associated with PFAS exposure, has raised public health concerns [9, 14]. At the end of their useful life most PFAS-containing products are disposed of in a landfill. While efforts have been made to reduce reliance on landfill disposal, landfills continue to be the most common municipal solid waste (MSW) management option in the United States (US).

Landfill leachate is an aqueous matrix formed when waste constituents (including PFAS) partition from the waste to the water present in landfills. Leachate collected by the leachate collection system can be derived from moisture within the solid waste, rainfall on the landfill, and sometimes groundwater intrusion into the landfill leachate collection system. MSW landfill leachate contains elevated concentrations of salts, organic matter, and ammonia generated during the decomposition of biodegradable wastes. Leachate also contains metals, PFAS, and other trace constituents originating from the waste [22,25,32,34].

To prevent these constituents from migrating into surrounding soil and groundwater, low permeability compacted soil (often clay) and geomembrane liners are used as a containment barrier. Nonetheless, the potential for chemical diffusion—particularly of organic compounds—through these liners has always been a concern [12,19]. Laboratory experiments and models have shown that chemicals can transport through geomembrane liners via both aqueous and gas phases or through water vapor, raising questions about the long-term containment capabilities of these barriers [38,39,40].

More recently, PFAS diffusion through these liners has been observed in a laboratory setting, with very low diffusion coefficients [11] and breakthrough time estimates of approximately 1500 years for intact liners [1]. In the field, however, studies have identified liner placement quality as a significant factor affecting constituent transmission rates as it contributes to the density of liner imperfections like tears and pinholes [19,41], and recent research has highlighted uncertainty in the effectiveness of liners against PFAS transmission in full-scale landfill operations [15]. Laboratory experiments to evaluate the transport of perfluooctanoic acid (PFOA) through liners under different conditions found that in most cases, a single composite liner barrier system is unlikely to contain PFOA to an acceptable level [3].

For further protection, some landfills in the US utilize a double-lined leachate collection system (LCS), designed with an upper liner and primary LCS as well as a lower liner and secondary LCS (also referred to as a leak detection system). Additional background information regarding landfill liner regulations at the US and Florida state-level are included in the supplementary information (SI) Section S1 and a cross-section of a typical double liner system is included in Fig. 1. In all cases, the underlying layer of low permeability (10⁻⁷ cm s⁻¹) soil/minerals further reduces leachate migration to groundwater and control groundwater moving into the LCS. Certain liner designs or improper construction combined with a shallow groundwater table can create an environment where the hydraulic head outside the landfill is higher than in the secondary leachate collection system, thus, a hydraulic gradient can drive groundwater into the secondary collection system [21,50].

Fig. 1.

A typical MSW landfill double liner leachate collection system (sawtooth liner), with major potential sources of secondary leachate.

Major potential sources of liquids to the secondary LCS are primary leachate and groundwater, as presented in Fig. 1 [43,50,4]. Constituents in the secondary leachate, such as PFAS, may originate from primary leachate passing through the geomembrane due to tears or pinholes, or by diffusion through the geomembrane itself [41]. Constituents may also be derived from other sources besides primary leachate, such as contaminated groundwater [6], landfill gas [29,48], or gas condensate, which, while a low volume input, contains many dissolved constituents [43], or from materials which are part of the leachate collection system itself [33].

Despite this, peer-reviewed research analyzing PFAS in landfill leachate collected from the secondary LCS has yet to be published [53]. This study focuses on the analysis of primary and secondary leachate from three Florida (USA) MSW landfills equipped with double liner systems to identify possible PFAS sources and assess the effectiveness of geomembrane liners in controlling PFAS migration. This study is the first to report PFAS concentrations in secondary MSW landfill leachate and provides a valuable discussion of potential migration pathways and PFAS partitioning within landfill leachate management systems.

2. Methodology

2.1. Site background and sample collection

Three active MSW landfills where the leachate collection systems included multiple sample access points for both primary and secondary leachate within multiple cells were identified, as presented in Table 1. All three sites have accepted MSW, which consists predominantly of household waste, yard waste, residential construction and demolition debris, and potentially some small fractions of industrial wastes like wastewater treatment biosolids, all waste streams which are likely to contain PFAS [2,8,10,23,26,31,46]. The climate data and rainfall for the three selected facilities were comparable (130–140 cm rainfall per year). Furthermore, the leachate collection systems at all three facilities were composed of two 60-mil high-density polyethylene (HDPE) bottom liners with high permeability geonets, as presented in Fig. 1. More information about the sites and sampling locations are included in the SI Section S2.

Table 1.

Sample location background information.

Facility and Region	Sampling Location	Cell Open	Cell Active at the Time of Sampling?	Footprint (ha)	Leachate Generation Rate (L/ha/day)
					Primary	Secondary
A – Southwest Florida	1	2010	Active	5.8	2200a	n/ae
	2	2010	Active	5.2	2200a	n/a
	3	2010	Active	5.2	2200a	n/a
	4	2010	Active	5.8	2200a	n/a
B – North Florida	5	2000	Closed in 2001	3.5	1300b	n/a
	6	2007	Closed in 2011	3.4	1300b	n/a
C – Central Florida	7	2007	Inactive	11.9	2200c	220c
	8	2007	Inactive	11.9	2200c	220c
	9	2017	Active	4.2	5300d	36d
	10	2017	Active	4.2	5300d	36d
	11	2017	Active	3.8	5300d	36d

^aleachate generation rates available for 2018 combined primary plus secondary leachate for four cells within one landfill construction phase.

^bleachate generation rates available for 2019 combined primary plus secondary leachate for six cells (including the two sampled in this study) within one landfill construction phase.

^cdiscrete leachate generation rates available for 2019 primary and secondary leachate for two cells in one landfill construction phase.

^ddiscrete leachate generation rates available for 2019 primary and secondary leachate for three cells in one landfill construction phase.

^ethis information was not available at the time of this study

MSW landfill leachate samples were collected during July (Facilities A and B) and August (Facility C) of 2020 from the primary and secondary leachate collection systems of three sites at a total of 11 sampling locations (22 samples). Sample collection was conducted over a single day for each site; additional sample collection methodology detail is included in the SI Section S2 Furthermore, background water concentration data from at least 20 unique groundwater monitoring wells available at each site was downloaded from an online portal [13] as presented in Table 2. Groundwater PFAS data for these sites from previous work by the authors [6] were used as groundwater background concentrations (SI Table S10).

Table 2.

Physical-chemical parameters among 11 primary (L_P) and 11 secondary (L_S) leachate samples collected from three Florida MSW landfills and mean ± standard deviation for available groundwater (GW) parameter data from the year of sampling.

Fac.	Loc.	pH			Ammonia-N (mg L−1)			Chloride (mg L−1)			Conductivity (μs cm−1)		COD (mg L−1)
		LP	LS	GW	LP	LS	GW	LP	LS	GW	LP	LS	LP	LS
A	1	7.3	6.4	6.4 ± 0.3 n = 40	820	30	7.8 ± 11 n = 41	2000	80	51 ± 45 n = 34	10,000	890	1600	980
	2	7.5	6.3		160	56		6800	80		16,000	830	4700	160
	3	7.2	6.8		57	320		4400	1100		10,000	7100	2600	770
	4	6.9	6.4		250	120		3200	40		6600	1500	1200	280
B	5	7.7	6.8	4.8 ± 0.7 n = 43	760	120	0.09 ± 0.3 n = 40	2400	400	23 ± 17 n = 40	8100	2000	2200	20
	6	6.5	6.5		140	51		400	200		1500	990	21	20
C	7	8.4	8.3	6.6 ± 0.5 n = 96	350	390	1.2 ± 2.1 n = 41	1600	1800	37 ± 52 n = 41	8600	9100	940	960
	8	8.6	9.5		590	860		2200	5000		13,000	21,000	1300	2800
	9	7.1	6.6		110	51		360	180		2700	1500	170	70
	10	7.1	6.7		38	0.20		220	60		2000	1100	230	40
	11	7.7	6.8		530	88		2800	60		19,000	950	2000	60
Average		7.5	7.0	6.1	350	190	3.1	2400	820	34	8.900	4300	1500	560

2.2. Sample analysis

Leachate samples were analyzed for physical-chemical properties (chloride, ammonia-nitrogen, pH, conductivity, total dissolved solids (TDS), total solids (TS), alkalinity, chemical oxygen demand (COD)), and metals according to methodologies listed in the SI Table S1. PFAS analysis of all samples was conducted according to a sample extraction process adapted from Robey et al. [36] and has been validated across multiple studies [31,43,37]. Details of the PFAS solid phase extraction protocol are included in the SI Section S3 (Table S2). The PFAS analytical method included 92 PFAS (full analyte list in SI Table S3 and calibration curve details in Table S4), and extracts were analyzed using a Thermo Scientific Vanquish ultra-high pressure liquid chromatograph (LC) coupled to a TSQ Quantis triple quadrupole mass spectrometer (UHPLC-MS/MS). Details about PFAS quality control are included in the SI section S4.

2.3. Data processing

When calculating summary statistics across samples, the concentrations of physical-chemical constituents (e.g., chloride, COD, ammonia-nitrogen) and trace metals below the detection limit were substituted with the value of their respective detection limit. The value of their respective detection limit was also substituted for PFAS which were below detection limits, and concentrations which are above detection limit but below quantitation limits are noted as semi-quantitative when reported. Only the PFAS compounds which were quantified in over 50 % of primary leachate samples were used for comparison between sample types and between predicted and observed secondary leachate concentrations. PFAS which met this criterion are listed in the results section (Tables 3 and 4).

Table 3.

Percent (%) of samples in which each PFAA was detected, median and maximum concentrations (ng L⁻¹, two significant figures), and average ratio between observed and expected secondary leachate PFAS concentrations in 11 primary and secondary leachate samples.

Class	Analyte	Method Detection Limit (ng L−1)	Primary Leachate (n = 11)			Secondary Leachate (n = 11) [Observed]			Secondary Leachate (n = 11) [Expected]
			DFa	Median	Max	DFa	Median	Max	Median	Max
PFCAs	PFBA	0.55	100 %	810	1700	100 %	390	1600	120	1900
	PFPeA	1.51	100 %	1600	2600	100 %	710	2200	56	3800
	PFHxA	0.49	100 %	2700	7900	100 %	1400	4700	120	10,000
	PFHpA	2.04	100 %	360	1200	100 %	230	880	10	1100
	ΣPFOA	2.23	100 %	610	5600	100 %	570	2800	20	3900
	PFNAlinear	1.23	91 %	21	210	100 %	13	190	1.2	100
	PFDA	0.30	91 %	20	650	100 %	25	1000	0.48	310
	PFUdA	0.74	45 %	< 0.74	16	45 %	< 0.74	48	n/a	n/a
	PFDoA	1.10	55 %	1.2*	14	36 %	< 1.1	75	1.1	15
	PFTrDA	0.86	18 %	< 0.86	2.9	27 %	< 0.86	9.9	n/a	n/a
	PFTeDA	1.00	36 %	< 1.0	6.7	45 %	< 1.0	53	n/a	n/a
PFSAs	PFPrSlinear	1.19	36 %	< 1.2	37	9 %	< 1.2	7.1	n/a	n/a
	PFBS	1.44	100 %	850	2300	91 %	67	4000	22	2500
	PFPeSlinear	2.87	27 %	< 2.9	1000	36 %	< 2.9	260	n/a	n/a
	PFHxSlinear	3.49	100 %	280	1200	82 %	23	580	24	660
	PFHpS	2.41	18 %	< 2.4	14	0 %	< 2.4	< 2.4	n/a	n/a
	ΣPFOS	2.30	91 %	25	152	100 %	18	250	3.0	58
	PFDS	1.28	9%	< 1.3	2.1	9%	< 1.3	6.1	n/a	n/a
PFSiAs	PFBSI	22.31	27 %	< 22	71	18 %	< 22	75	n/a	n/a

n/a: not applicable, this compound was detected in less than 50 % of samples

^*reported value greater than detection limit, less than limit of quantitation

^aDF = detection frequency

Table 4.

Percent (%) of samples in which each precursor PFAS was detected, median and maximum concentrations (ng L⁻¹, two significant figures), and average ratio between observed and expected secondary leachate PFAS concentrations in 11 primary and secondary leachate samples.

Class	Analyte	Method Detection Limit (ng L−1)	Primary Leachate (n = 11)			Secondary Leachate (n = 11) [Observed]			Secondary Leachate (n = 11) [Expected]
			DFa	Median	Max	DFa	Median	Max	Median	Max
FASAs	FOSA	1.47	64 %	2.2*	12	45 %	< 1.5	17	1.5	20
	FOSAA	0.66	36 %	< 0.66	15	36 %	< 0.66	83	n/a	n/a
	N-MeFOSAA	0.68	91 %	32	350	64 %	4.5	1300	2.2	160
	N-EtFOSAA	0.78	91 %	7.7	93	45 %	< 0.78	530	n/a	n/a
	N-MeFOSE-M	1.41	55 %	2.7*	67	36 %	< 1.4	62	0.66	9.3
	N-EtFOSE-M	1.41	36 %	< 1.4	39	27 %	< 1.4	17	n/a	n/a
	N-AP-FHxSA	1.20	9 %	< 1.2	13	27 %	< 1.2	8.7	n/a	n/a
	N-CMAmP–6:2 FOSA	1.72	64 %	8.0	20	18 %	< 1.7	57	1.7	18
Fluorotelomer PFAS	6:2 FTS	19.67	100 %	110	290	55 %	33*	350	20	250
	8:2 FTS	19.55	18 %	< 20	120	18 %	< 20	190	n/a	n/a
	10:2 FTS	0.99	36 %	< 0.99	9.7	18 %	< 0.99	94	n/a	n/a
	6:2 FTCA	1.01	100 %	2600	4700	100 %	2000	14,000	120	1800
	8:2 FTCA	1.23	100 %	40	1000	82 %	30	1800	9.1	480
	10:2 FTCA	0.49	55 %	7.0	47	45 %	< 0.49	330	0.49	20
	8:2 FTUCA	2.99	55 %	9.6	46	64 %	15	670	2.3	42
	10:2 FTUCA	0.62	9 %	< 0.62	3.3	45 %	< 0.62	48	n/a	n/a
	3:3 FTCA	2.73	100 %	250	880	100 %	320	2900	60	920
	5:3 FTCA	4.82	100 %	7500	50,000	100 %	2400	57,000	1300	17,000
	6:3 FTCA	3.21	91 %	85	500	82 %	170	820	7.4	480
	7:3 FTCA	3.38	100 %	310	4600	91 %	390	4900	67	2400
	8:3 FTCA	2.56	55 %	7.0*	19	64 %	19	81	2.6	17
	5:2sFTOH	1.98	100 %	2700	3900	82 %	1800	11,000	92	1600
	7:2sFTOH	2.47	55 %	89	1000	45 %	< 2.5	1800	2.8	470
Other Precursor PFAS	6:2 diPAP	0.68	100 %	71	130	55 %	4.5	29	12	68
	Syn40	12.70	27 %	< 13	1100	9 %	< 13	1500	n/a	n/a
	Syn41	4.25	100 %	180	420	55 %	25	490	12	330
	Syn45	7.71	64 %	400	580	0 %	< 7.7	< 7.7	< 7.7	920
	Syn53	9.41	73 %	220	1500	55 %	36	1200	11	920
	Oak6	0.70	18 %	< 0.70	59	45 %	< 0.70	41	n/a	n/a
	Oak8	3.52	82 %	100	280	36 %	< 3.5	260	3.5	240
	Oak10	3.07	36 %	< 3.1	8.4	27 %	< 3.1	9.1	n/a	n/a

n/a: not applicable, this compound was detected in less than 50 % of samples

^*reported value greater than detection limit, less than limit of quantitation

^aDF = detection frequency

Primary and secondary leachate data sets were tested for normality using a Jarque-Bera test. Non-parametric data were transformed to achieve normality, if possible, and the original data (if normally distributed) or transformed data underwent a paired one-tail two-sample student’s T-test to identify statistically significant differences. This statistical analysis method was selected to compare pairs of data representing leachate in one cell above and below the primary LCS liner, with a hypothesis that secondary leachate concentrations will be lower. Non-parametric data which could not be transformed to achieve normality underwent a Wilcoxon signed-rank test. For parametric, transformed, and non-parametric analyses, a p-value less than 0.05 indicates a statistically significant difference in a given parameter between primary and secondary leachates from the same cells.

Groundwater (background) concentrations of chloride (see Table 2), a conserved constituent, were used to estimate the proportion of secondary leachate [50] which is derived from primary leachate (${\mathrm{\%}}_{\mathrm{P}}$) vs groundwater (assuming other possible sources contribute de minimis volumes) using Eq. 1. Estimated fractions of secondary leachate sources (${\mathrm{\%}}_{\mathrm{P}}$) based on conserved constituents may, in turn, be used to calculate expected concentrations of PFAS in secondary leachate (using Eq. 2, solving for an expected $C_S$), which may then be compared to the observed concentrations in secondary leachate. Observed concentrations which are significantly higher or lower than expected may indicate unknown PFAS sources or partitioning behavior within the secondary LCS.

$${\mathrm{\%}}_P=\frac{C_S-C_{GW}}{C_P-C_{GW}}$$ (1)

$$C_S=\left(C_P\mathrm{*}{\mathrm{\%}}_P\right)+\left(C_{GW}\mathrm{*}\left(1-{\mathrm{\%}}_P\right)\right)$$ (2)

Where, ${\mathrm{C}}_{\mathrm{S}}$ = the constituent concentration in secondary leachate (m/V), ${\mathrm{C}}_{\mathrm{P}}$ = the constituent concentration in primary leachate (m/V), ${\mathrm{\%}}_{\mathrm{P}}$ = the percent (between 0 and 1) of secondary leachate which is derived from primary leachate, and ${\mathrm{C}}_{\mathrm{G}\mathrm{W}}$ = the constituent concentration in groundwater (m/V).

Chloride-normalized PFAS concentrations were calculated as presented in Eq. 3 [50].

$$PFA{S}_S=\frac{C{l}_S\mathrm{*}PFA{S}_P}{C{l}_P}$$ (3)

Where, $\mathrm{P}\mathrm{F}\mathrm{A}{\mathrm{S}}_{\mathrm{S}}$ = Expected PFAS concentration in the secondary leachate (m/V), $\mathrm{P}\mathrm{F}\mathrm{A}{\mathrm{S}}_{\mathrm{P}}$ = PFAS concentration in the primary leachate (m/V), $\mathrm{C}{\mathrm{l}}_{\mathrm{P}}$ = chloride concentration in the primary leachate (m/V), and $\mathrm{C}{\mathrm{l}}_{\mathrm{S}}$ = chloride concentration in the secondary leachate (m/V).

3. Results and discussion

3.1. Physical-chemical parameters

Variability in MSW landfill leachate constituent concentrations is well-documented [22], and while physical-chemical parameters and metal concentrations in the 22 samples ranged by one to two orders of magnitude, the ranges were typical for MSW landfill leachate, as presented in Table 2 and in the SI Section S5, Fig. 2 and S1 (for physical-chemical parameters), and Table 2, S7, S8, and S9 (for metals) [22,34]. The primary leachate pH ranged from 6.5 to 8.6, ammonia from 38 to 820, chloride from 220 to 6800, COD from 21 to 4700 mg L⁻¹, and conductivity from 1500 to 19,000 μS cm⁻¹as presented in Table 2. The total metals concentrations in the primary leachates ranged from 100 to 2300 mg L⁻¹.

Fig. 2.

Chloride, COD, ammonia-nitrogen, and total metals concentrations (mg L⁻¹) in 11 samples of primary and secondary leachate, and p-values from a one-tailed paired sample t-test. The box-and-whisker plots represent the 10th, 25th, 50th, 75th, and 90th percentile, “X” indicates the mean.

The secondary leachate pH ranged from 6.3 to 9.5, ammonia-nitrogen from 0.2 to 860, chloride from 40 to 5000, COD from 20 to 2800 mg L⁻¹, and conductivity from 830 to 21,000 μS cm⁻¹ as presented in Table 2. Total metals concentrations in the secondary leachates ranged from 36 to 2700 mg L⁻¹. Except for ammonia-nitrogen (p = 0.175), all physical-chemical parameters, including total metals, were statistically significantly lower (p < 0.01) in the secondary than primary leachates samples as presented in Fig. 2. Differences between constituent concentrations in the primary and secondary leachates may be a result of leachate variability, evaporation due to longer retention times in the secondary leachate collection system resulting in higher concentrations, or dilution from groundwater intrusion resulting in lower concentrations.

3.2. PFAS profiles in leachate

3.2.1. Primary leachate

The PFAS content in primary landfill leachate is predominantly derived via direct leaching from the waste mass, with leached PFAS concentrations influenced by multiple factors, including waste composition, age, and climate [25,44,49]. Of the 92 PFAS included in the analytical method of this study, a total of 50 compounds (Σ₅₀PFAS) were quantified in at least one leachate sample. The number of unique PFAS measured in individual samples ranged from 19 to 43 (mean: 33 compounds) per sample. Among the detected PFAS, the fluorotelomer carboxylic acid 5:3 FTCA was the most abundant in nine of the 11 samples, while 6:2 FTCA and PFOA were the most abundant in the remaining two samples. Detection frequency (DF), median, and maximum concentrations of PFAS across all samples are presented in Table 3 (for terminal PFAS (PFAAs)) and Table 4 (for precursors and other PFAS) and their distribution is shown in Figure S3. Terminal PFAS (Σ₁₉PFAAs) represented 29 %, on average, (median of 21 %) of Σ₅₀PFAS in primary leachate while precursors represent the remainder.

Total PFAS concentrations (Σ₅₀PFAS) among primary leachate samples in this study were, on average, higher than those previously reported for MSW landfill leachate in the US [6,49], ranging from 3200 to 81,000 ng L⁻¹ (mean: 29,000 and median: 23,000 ng L⁻¹) as presented in Table 5. Higher concentrations may be a result of the expanded suite of compounds included in the analytical method since at the individual PFAS level, the PFAS concentrations quantified (e.g., PFOA, PFOS) were similar to those previously reported [49].

Table 5.

PFAS Concentrations (ng L⁻¹) in primary and secondary leachate.

Sample ID	ΣPFOAa		ΣPFOS		PFBA		PFHxA		PFNAlinear		PFUnDA		PFDoDA		PFTeDA		PFBS		PFHxS		ΣPFASb			# PFAS
	LP	LS	LP	LS	LP	LS	LP	LS	LP	LS	LP	LS	LP	LS	LP	LS	LP	LS	LP	LS	LP	LS	LP	LS
1	1300	660	27	8.6	1200	450	7900	2200	62	21	1.6 *	< 0.74	< 1.1	< 1.1	3.5	3.8	940	19	350	4.8 *	81,000	32,000	37	25
2	640	42	28	7.4	1300	60	3200	140	22	3.8	2.7	< 0.74	4.7	< 1.1	4.3	< 1.0	2300	30	350	< 3.5	56,000	28,000	43	24
3	230	240	11	18	850	940	1200	1400	3.7 *	2.2 *	< 0.74	< 0.74	1.2 *	< 1.1	< 1.0	< 1.0	240	130	280	130	28,000	96,000	32	29
4	100	110	15	54	610	67	830	280	2.1 *	13	< 0.74	3.5	< 1.1	< 1.1	< 1.0	< 1.0	120	< 1.4	120	< 3.5	15,000	3300	27	23
5	1900	2600	150	250	1700	390	3700	1400	67	190	2.7	48.4	4.6	75	3.1	< 1.0	370	80	1200	170	37,000	16,000	42	42
6	5600	1700	74	25	811	200	5200	810	210	150	15.5	45.6	11	42	< 1.0	3.0	1100	67	61	35	34,000	11,000a	41	38
7	610	2600	20	58	570	880	2700	3700	21	68	7.4	3.8	14	6.7	< 1.0	< 1.0	850	1000	120	580	14,000	28,000	36	34
8	1700	2900	25	53	810	1600	4600	4700	34	78	< 0.74	3.4	1.3 *	5.3	< 1.0	< 1.0	1100	4000	290	470	23,000	39,000	32	34
9	3.8 *	150	< 2.3	5.4 *	280	490	130	1400	< 1.2	3.3 *	< 0.74	< 0.74	< 1.1	< 1.1	6.7	3.9	22	42	5.2 *	7.7 *	12,000	9700	19	18
10	160	580	49	17	110	280	200	1200	6.1	9.7	< 0.74	< 0.74	< 1.1	< 1.1	< 1.0	52.8	170	76	190	23	3200	8500	26	18
11	190	30	20	9.6	480	42	840	160	3.4	1.7 *	< 0.74	< 0.74	0.32	< 1.1	< 1.0	10.6	870	21	280	8.4 *	21,000	11,000	25	20
RSLc	6		4		1800		990		5.9		600		100		2000		600		39
MCLd	4		4						10										10

Sample concentrations in bold text exceed their respective RSL (THQ=0.1), non-detects are reported as less than the limit of detection, and values greater than the limit of detection but less than the quantitation limit are indicated with an asterisk (*).

^aOnly PFAS which have been evaluated for toxicity and assigned regulatory or risk-based thresholds and which were quantified in this study are included

^bAll samples except for L_S collected from sample location 6 included at least one PFAS measured above the detection limit but below the limit of quantitation, included in the summation of total PFAS.

^cUS EPA Regional Screening Levels (RSL), risk-based screening levels for Superfund sites

^dUS EPA Maximum Contaminant Limits (MCL), enforceable primary drinking water standards

3.2.2. Secondary leachate

Surprisingly, given the significantly lower concentrations of most physical-chemical parameters, Σ₅₀PFAS concentrations in secondary leachate samples were also higher, on average, than those reported for primary MSW landfill leachate in previous studies [6,25,49]. The detection Frequency (DF), median, and maximum concentrations of PFAS across all samples of secondary leachate are presented in Table 3 (PFAAs) and Table 4 (precursor and other PFAS). In total, 48 PFAS were quantified in secondary leachate, with individual samples containing between 18 and 42 PFAS (mean: 28). The most abundant PFAS in six of the 11 secondary leachate samples was 5:3 FTCA. In three samples, 6:2 FTCA was the most abundant, while 7:2 saturated fluorotelomer alcohol (7:2 sFTOH) and PFOA were most abundant in the remaining two samples. The Σ₁₉PFAA represented, on average, 26 % (median: 30 %) of the total PFAS in secondary leachate, with an average Σ₁₉PFAAs concentration of 5200 ng L⁻¹ (median: 4100 ng L⁻¹) as presented in Fig. 3.

Fig. 3.

PFOA, PFOS, and total PFAS concentrations in 11 samples of primary and secondary leachate, and p-values from a one-tailed paired sample. The box-and-whisker plots represent the 10th, 25th, 50th, 75th, and 90th percentile, “X” indicates the mean.

3.2.3. PFAS in primary vs. secondary leachate

Unlike most of the physical-chemical parameters, the Σ₅₀PFAS concentrations in secondary leachate were, on average, 13 % lower but not significantly different than those of the primary leachate (p = 0.2). Furthermore, the fraction of terminal PFAS in secondary leachate was not significantly different from primary leachate (p = 0.1) as presented in Fig. 3.

On a compound-specific basis, concentrations of two precursor PFAS, 6:2 diPAP and Syn45 (dodecafluorosuberic acid, a perfluoroalkyl dicarboxylic acid), were significantly different between primary and secondary leachate (p < 0.003). Both compounds were present at lower concentrations, or not detected at all, in secondary leachate. Syn45, detected in seven primary leachate samples at an average concentration of 480 ng L⁻¹, was not detected in any secondary leachate samples. Additionally, 2H-perfluoro-2-dodecenoic acid (10:2 FTuCA) was detected in only one primary leachate sample but present in five secondary leachate samples. This may indicate that it is an intermediate transformation product, likely resulting from the breakdown of precursors such as diPAPs in the secondary leachate collection system [27,45,46], or it may originate from a different source. Additional comparison between primary and secondary leachates is included in the SI Section S6 (Tables S10 and S11, Figures S2 through S6).

Notably, although ΣPFAS in this study was higher, on average, than previous studies, the average concentration of all PFAS for which average US MSW landfill leachate concentrations were reported by Tolaymat et al. [49] – PFOA, PFOS, PFNA, PFBS, PFHxS, and PFHxA – were lower in this study than in the review. This may be due to the phase-out of PFOA and PFOS over the past 20 years once again underscores the significance of precursor PFAS in landfill leachate, which has not been adequately addressed in historical analyses that mainly focused on terminal PFAS, such as those listed in Table 5.

3.3. PFAS attribution in secondary leachate

Lower pH in secondary leachate compared to primary leachate, and the lower pH observed in groundwater, may indicate groundwater infiltration lowering the pH in the secondary leachate collection system (Table 2 and S10). Similar trends for ammonia-nitrogen and chloride concentrations—where groundwater levels are lower than those in primary leachate—further support the possibility of groundwater impacting secondary leachate composition. Based on chloride concentrations used as a stable tracer to estimate the contribution of groundwater and primary leachate, secondary leachate was estimated to contain, on average, 33 % primary leachate and 67 % groundwater (Eq. 2, Section 2.3; Table S11). In cases where secondary leachate chloride concentrations exceeded those in primary leachate, secondary leachate was assumed to contain 100 % primary leachate.

PFAS concentrations in groundwater samples collected from these landfills as part of previous studies [6] are included in Table S10. These concentrations represented, on average, 0.3 % of the total PFAS in the secondary leachate. Thus, while groundwater contributes significantly to secondary leachate volume in several sample locations, PFAS present in groundwater was unlikely to contribute significant PFAS concentrations to the secondary leachate. This suggests that the PFAS in the secondary leachate are mainly derived from PFAS from within the landfill itself.

Assuming primary leachate is the sole PFAS source contributing to secondary leachate concentrations, the expected PFAS levels in secondary leachate were calculated using Eq. 3 and are included in Table 3 (PFAAs) and Table 4 (precursor and other PFAS). The observed PFAS concentrations were significantly higher than those calculated, with the observed-to-expected ratios ranging from 0.7 to 117, averaging 26 times higher. On an individual compound level, approximately 57 % of data points had ratios above 1.0, indicating higher-than-expected PFAS concentrations in the secondary leachate, while 22 % of the data showed lower-than-expected levels, and 20 % could not be assessed due to detection limits. Terminal PFAS were more likely to exceed expected levels, with 74 % of the PFAA data points exceeding expected concentrations, compared to 48 % for precursors, suggesting additional factors influencing PFAS behavior in the system. The distribution of observed: predicted secondary leachate concentrations for all PFAS is included in Fig. 4.

Fig. 4.

Distribution of ratios between predicted and observed secondary leachate PFAS concentrations for (top) all PFAS and PFAS with detection frequency (DF) greater than 50 %, and (bottom) split between terminal and precursor PFAS. Data points where both predicted and observed concentrations are below the limit of detection are excluded.

The effectiveness of geomembranes in controlling the movement of organic compounds varies depending on both the geomembrane and organic compound of interest’s structures. Similarly, PFAS interaction with landfill liner systems and their transmission from the primary to the secondary leachate collection system may be influenced by individual chemical structures, such as PFAS class and chain length. Shorter-chain terminal PFAS have lower molecular weights and are more polar than longer-chain and precursor PFAS, making it unlikely for terminal PFAS to pass through geomembrane liners via diffusion alone.

Both the general trends related to concentrations of select PFAS in primary and secondary leachates, as well as the presence of intermediate transformation products in the secondary leachate that are not present in primary leachate indicate that PFAS profiles in the secondary leachate are not solely a result of PFAS migration via liner imperfections, and may, instead, result from other factors such as PFAS sorption and transformation [54]. Conditions within the secondary leachate collection system differ from the primary leachate collection systems in a number of ways that support this hypothesis. Secondary leachate collection systems are designed to detect leakage from the primary leachate collection system, so typical secondary leachate generation rates are much lower than primary leachate, resulting in longer retention times, which means more opportunity for PFAS transformation and interaction with the liner (i.e., sorption). as will be discussed later in the manuscript.

Additional time for transformation would result in higher concentrations of intermediate transformation products and terminal PFAS – both of which were disproportionately higher in the secondary leachates in this study. Over time, large molecular weight precursor PFAS transform into smaller intermediate PFAS and eventually PFAAs. This, alone, would impact PFAS profiles in the secondary leachate, but differences in partitioning behavior (e.g., into the gas or onto the solids) among different classes and sizes of PFAS also influence leachate-liner interaction and the likelihood for PFAS to sorb or remain mobile within the secondary leachate collection system [35,42]. Larger molecular weight PFAS (e.g., long-chain PFAAs and precursors) tend to preferentially sorb onto liners and solids (e.g., HDPE liners) compared to short-chain PFAAs [5,33]. This may explain the higher ratio of shorter-chain PFAS in secondary leachate (vs. expected concentrations) compared to the longer-chain. Notably, ultra-short-chain PFAS like PFPrA, which have been reported in landfill liners, were not included in this study [33].

Landfill gas migrating to the secondary system may also contribute to the elevated PFAS levels in secondary leachate. Compounds such as 7:2 FTOH, commonly found in landfill gas, and corresponding intermediate transformation products like FTCAs could enter the leachate system and significantly raise PFAS concentrations in the secondary leachate [29, 43,55]. Leachate collected above the first liner also comes in contact with landfill gas, however, longer retention times and greater leachate surface area-to-volume ratio in the secondary leachate collection system mean that PFAS from landfill gas have more opportunity to partition to the secondary leachate.

Given the lack of direct empirical evidence, this analysis relies on documented mechanisms that could influence PFAS concentrations in landfill environments and leachate. While diffusion through the geomembrane is theoretically possible, it is more likely that factors such as residence time, sorption, and transformation processes within the landfill system are the main contributors to the observed PFAS concentrations. Specifically, the disproportionately high concentrations of select PFAS, especially terminal PFAS, are likely a result of a combination of PFAS sorption, transformation, landfill gas, and other contributions specific to each landfill’s leachate collection system design and operation.

3.4. Limitations

This study reports PFAS and physical-chemical parameter data for single grab samples of leachate collected from full-scale MSW landfill LCS. As such, they may not necessarily be representative of leachate from their respective landfills or reflect variability within or across sites as a result of changes in weather, precipitation, or waste composition. Collecting samples from 11 landfill cells across three landfills representing different waste ages does provide some reflection of the potential variability one might expect to see across repeated sampling, but leachate quality data for replicate sample collection events would offer additional value. In a similar vein, all of the sites visited for this study were in Florida, with similar climates and precipitation, which presents another limitation to the study. Additional sample data from a greater variety of sites and sources will always provide a greater understanding of not only the variability of leachate as a matrix, but different sites and conditions may also shed additional light on the mechanisms of transport and transformation proposed here.

Processes taking place under complex, out-of-sight conditions within a full-scale MSW landfill leachate collection system will always be limited to hypotheses and proposed mechanisms. Laboratory-scale experiments under controlled environments will provide a better and more precise understanding of PFAS behavior in these conditions. This study suggests possible mechanisms which should be included in future studies (i.e., sorption and other interaction with LCS materials, and the role of retention time in PFAS transformation under anaerobic conditions).

3.5. Implications and conclusions

In this study, PFAS were detected in all primary and secondary leachate samples, indicating their transport to secondary leachate collection systems via multiple proposed mechanisms. The disproportionately high concentrations of many PFAS indicates that conditions within the secondary LCS itself may be contributing to increased PFAS transformation, or that PFAS may be derived from another source, such as landfill gas. It is most likely that the observations are a result of a combination of mechanisms. This is the first report of PFAS concentrations in secondary leachate and provides important guidance for future studies to more closely evaluate these mechanisms under both a controlled laboratory environment and through full scale field sampling.

Although all three MSW landfills in this study use double HDPE geomembrane liner systems, most US landfills use a single liner. The average performance efficiency of landfill liners in the US is estimated to be approximately 98 %, with a calculated theoretical PFAS emissions of 14 kg per year [21,49]. Studies have shown that PFAS can pass through non-membrane, low-permeability landfill liners, such as compacted clay [28], so liner placement, integrity, and proper contact between the bottom liner and the low-permeability soil or clay layer are crucial for reducing leachate and associated PFAS release to the environment [17, 16,3]. It is noted that direct emissions through secondary liners, such as the ones in this study, are likely lower than transmission through primary liners, due to the lower hydraulic pressure on the secondary liner and the presence of the underlying clay layer. Also, the observation that many of the secondary leachates contained proportionally more groundwater than primary leachate indicates that several of these sites were experiencing groundwater intrusion, not leachate emission, through the bottom liner.

The US Environmental Protection Agency (US EPA) has set risk-based guidelines (tap water regional screening levels; RSLs, and enforceable drinking water maximum contaminant limits MCLs) for 13 PFAS, including 12 PFAAs and the replacement PFAS GenX [51,52]. Out of the 12 PFAAs, nine were detected in both primary and secondary leachate samples, while GenX was not included in the study’s analytical method. Table 5 shows a comparison of the concentrations of these nine PFAS across the 22 leachate samples to their respective RSLs (THQ=0.1). As observed in previous studies of landfill leachates, the concentrations of PFOA were the highest in both primary and secondary leachates compared to its corresponding US EPA RSL. PFOA concentrations were approximately 188 times higher than the RSL for primary leachate and 176 times higher for secondary leachate. For reference, a 2023 review of leachate studies across the US reported PFOA concentrations in MSW landfill leachate to be higher than the RSL (THQ = 0.1) by a factor of 230 [49].

Possible PFAS emissions through landfill liners may pose a long-term environmental challenge. The findings of this study suggest that conditions within the secondary leachate collection system may contribute to higher-than-expected PFAS concentrations in secondary leachate compared to conserved constituents like chloride. Furthermore, previous studies which have reported PFAS in groundwater and the detection of PFAS in all secondary leachate samples in this study indicates some level of PFAS migration through landfill liners, highlighting the need for diligent leachate management, landfill liner quality control, and further research into PFAS-liner interactions—particularly concerning precursor PFAS.

Supplementary Material

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi: 10.1016/j.jhazmat.2025.138705.

HIGHLIGHTS.

• Leachate collected from primary and secondary leachate collection systems of three MSW landfills.

• Most secondary leachates had lower physical-chemical constituent concentrations than primary.

• Low chloride concentrations in the secondary indicate groundwater mixing with primary leachate.

• Secondary leachate PFAS was disproportionally high based on chloride concentrations.

• Transformation, liner interaction, and other PFAS sources are explored as possible explanations.

Data availability

Data will be made available on request.