Legacy 1 , 2 , 3-trichloropropane contamination : a systematic review of treatments

1,2,3-Trichloropropane (TCP), a suspected human carcinogen, is a widespread contaminant that leaches into groundwater, where it persists. This systematic review of studies examines treatment technologies for TCP contamination. A four-database search yielded 1,160 papers, 36 of which met the eligibility criteria for a full-text review. The three most-represented treatment technologies, such as biodegradation (13), zerovalent transition metals (8), and granular activated carbon (GAC) (4), are either fully deployed in water systems or in the field test stage. To meet TCP treatment goals, additional site-specific testing of well water is needed since source water chemistry and cocontamination influence treatment efficacy. Future studies should include standardized units for reporting degradation or sorption normalized to surface area, chemical input, and/or energy expenditures. Although GAC is the most common treatment for contaminated wells, this technology remains limited due to a low TCP adsorption capacity which requires frequent bed-volume replacement. Aerobic biodegradation, reduction with zerovalent iron, and Fenton’s treatment produce byproducts that could limit their use. A geospatial analysis of TCP treatment studies reveals a dearth of knowledge about the extent of TCP contamination. TCP contamination is documented in at least nine countries on three continents, but there is little information about the rest of the world.

As new chemicals are produced and disposed of without consideration for their ultimate fate in the environment, the chances increase that some of these chemicals will cause serious problems for global water quality and public health (Damania et al. ). Moreover, water and wastewater treatment plants are not always designed to treat these pollutants, and many will have to take a more costly and reactive 'end of pipe' approach to treatment (Alpizar et al. TCP contamination is widespread and detected in groundwater in North America, Europe, and Asia (Kielhorn et al. ). In 2011, the European Chemicals Agency put TCP on its Candidate List as a substance of very high concern (SVHC) because it is carcinogenic and toxic to reproduction (ECHA ). Although there is no U.S. federal maximum contaminant level (MCL), the allowable amount of a contaminant in drinking water delivered to consumers, TCP was listed on the U.S. Environmental Protection water systems, and/or dilute contaminated water to concentrations below the MCL (SWRCB ).
TCP can also be categorized as a 'contaminant of emerging concern', a legacy chemical with newly understood environmental and/or public health consequences (Sauve & Desrosiers ). The California SWRCB says that acute exposure to TCP can burn the skin and eyes and that breathing TCP can irritate the throat and lungs and affect concentration, memory, and muscle coordination. Long-term exposure in drinking water may damage the liver and kidneys and increase the likelihood of tumors in multiple organs (US EPA ). TCP ingestion has been shown to cause cancer in animals and is believed to be a cancer risk for humans (WHO ;

SWRCB ).
As of 2019, there has not been a systematic review of TCP groundwater treatment studies. Two earlier reviews of TCP treatment options were not systematic (Samin & Janssen ; Merrill et al. ). Uncertainty remains about how to proceed in the face of expensive monitoring and removal of TCP from public and private wells. The primary goal of this study is to determine the most effective TCP treatment method through a systematic review of the peer-reviewed research. To achieve this goal, this review will: (1) identify which groundwater treatments are being researched for TCP (both field and bench studies); (2) determine how contaminate reduction varies among treatment types; (3) determine chemical byproducts of the different treatments; (4) map research locations and TCP contamination sites; and (5) identify opportunities for future research.

This systematic review follows the Preferred Reporting
Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for systematic reviews (Moher et al. ). Included in this review are published nonsystematic reviews and research on TCP with special attention to treatment and/or remediation of contaminated water sources.
An initial search exposed a variety of studies ranging from the carcinogenic potential and toxicology of TCP to its use as a precursor to epichlorohydrin manufacture (van Leeuwen et al. ). Four online databases were consulted for this review: Science Direct, Web of Science, PubMed, and Engineering Village. The publication date was not considered as an exclusion factor.
From the total number of results in an initial search, papers were transferred to the citation manager, RefWorks, and de-duplicated manually by two reviewers. To confirm the results, the reviewers manually deleted duplicates after sorting by title, and once again after sorting by author.
Although the automatic de-duplicating function within RefWorks has been found to yield the smallest number of false positives (duplicate citations deleted in error) when compared with Mendeley and Endnote de-duplication functions, manual de-duplication produces the fewest deduplication errors (Kwon et al. ). After duplicates were removed, two reviewers screened the remaining abstracts, keywords, and titles for the presence of '1,2,3-trichloropropane' and direct reference to any variety of treatment to remove TCP from contaminated water. Research papers were included whether the treatment technology studied was in early bench-scale experimental stages or in fullscale use by water utilities.
Finally, the first author independently read the 36 fulllength unique studies and classified each by the treatment method, the percent of contaminate reduction or removal rate, and the location of each paper's primary author.
Additional results, identified through the references of selected papers, were added if they met the eligibility requirements above.

RESULTS AND DISCUSSION
Using 1,2,3-TCP as the single search term in each database produced a total of 1,160 results ( Figure 1); and four papers identified in references found in the other eligible papers were also included. All duplicates, a total of 199, were removed. The remaining 965 articles were screened using abstracts and titles, leaving 40 articles that meet the eligibility requirements (inclusion of 1,2,3-TCP and one or more treatment technology). Four papers were excluded because their full text was unavailable (Supplementary Material, Table S11). Overall, this study includes full-text reviews of 36 papers (Table 1).

Spatial distribution and temporal trends in TCP treatment research
In a 2003 report from the World Health Organization, TCP contamination in the hydrosphere was reported in Europe, North America, and Asia (Kielhorn et al. ). Globally, TCP concentrations in the hydrosphere range from its detection limit (5 ng/L) to >100,000 ng/L in both ground and surface waters (Table 2 and Figure 2). An objective of this study is to reveal spatial trends in TCP research by mapping where treatment studies are taking place in relation to contamination sites (Figure 2).
Sixty-five percent (23/36) of papers in this study analyze research conducted in the U.S. followed by the Netherlands (14%, 5 papers) ( Table 3). The U.S. and the Netherlands have significant groundwater contamination with TCP connected to agricultural application to orchards, vineyards, and potato crops (Samin & Janssen ; Babcock et al. ). Though none of the TCP research papers was published in South America, Africa, or Australia, the global  There appears to be an increasing interest in TCP in recent years. In the 5 years between 1990 and 1994, only one TCP treatment study was produced. In contrast, there were 11 studies published in the past 5 years ( Figure 3).

Treatment methods
Of the 36 TCP treatment studies considered for full-text review, 32 were experimental peer-reviewed research papers which fall into two categories: (1) separation-based technologies and (2) degradation-based technologies.
The largest number of studies (13, 36%) investigate       reduce TCP to meet regulatory levels, even at low initial concentrations (<2 μg/L), and also showed that reduction continued in the field for 15 months post-injection. That study shows that degradation rates are slower, the lower the TCP concentration. Higher inoculum concentrations are needed for reduction, and optimal reduction occurs in a pH range from 7 to 9 (Schmitt et al. ).

Zerovalent iron and zinc
Eight papers investigated zerovalent metals as reductants, namely iron and zinc to treat water contaminated with TCP. Zerovalent zinc (ZVZ) is a more potent reductant than zerovalent iron (ZVI), reducing TCP at rates one to three orders of magnitude faster (Salter-Blanc et al. ).
ZVZ successfully degrades halogenated alkanes, which includes TCP, of various sizes (Tratnyek et al. a). ZVZ fully reduces TCP to propane which avoids the accumulation of partially reduced products, but when treated with

).
In multiple bench-scale experiments using ZVZ, TCP was removed below detection limits (5 ng/L) for samples ranging in concentration from 30 to 10 5 micromolar (μM).
Kinetics experiments reveal that rate constants normalized for surface area (K SA ) were between 10 À3 and 10 À2 (L g À1 GACs are emptied, reactivated, or replaced when breakthrough concentrations are between 10 and 50 ng/L. Harada's () data also shows that GAC particle size affects performancethe smaller the particle size (mesh size 170 × 200), the quicker contaminants break through the filtration compared with larger sizes (mesh size 100 × 120). However, particle size performance may be due to the accuracy of the scaling equations used (Harada ).
Before the effects of GAC particle size can be fully understood, more information is required.

Persulfate oxidation
This review identified four studies that assess activated per- Early et al. () explained how changing specific parameters of the kinetics experiments affects the production of byproducts. Their study varied the amounts of C, Pt, and Sn exposed to a reaction mixture of TCP (3,000 ppm) and H 2 (15,000 ppm). Dechlorination products included propane, propene, allyl chloride, and dichloropropene. Catalysts with Pt:Sn ratios of 9:1 and 6:1 exhibited a higher relative hydrogenation activity than monometallic Pt/C. Though Pt/Sn C catalyzed hydrogenation reduced TCP, the Early et al. () study did not establish reaction rates nor removal percentages. In short, the feasibility of Pt/Sn C catalyzed hydrogenation as a treatment technology and for limiting the production of undesirable byproducts has yet to be determined.

TCP reviews, reports and treatment overviews
This review found one report (

Identification of TCP treatments
This study aimed to determine the most effective TCP treatment and/or remediation methods. An objective was to identify the types of groundwater treatments for TCP being researched. Treatments can be broadly divided into separation-based and elimination-based technologies. There is one ex situ separation-based technology: GAC, and eight elimination-based technologies: (1) bioremediation, (2) zerovalent zinc, (3) zerovalent iron, (4) persulfate oxidation, (5) Fenton's treatment, (6) ammonia treatment, (7) hydrogenassisted dechlorination, and (8) sonolysis.
Because degradation parameters are unique to a specific treatment technology (

Scale of testing and summary of removal information
Another objective of this review is to determine which TCP remediation methods have been field tested with TCP-contaminated groundwater; those limited to bench tests with water spiked with TCP; those that have undergone pilot testing; and those that are in full-scale operation (Figure 4 and Since ZVZ is a powerful oxidant, the occurrence of partially   Site-specific testing is necessary to establish co-contaminates and water chemistry for optimal GAC selection to decrease the operational costs of using this technology.
Although no information about TCP contamination was found in South America, Australia, and Africa, there is evidence of widespread global contamination of TCP in the hydrosphere. A complete global map of TCP contamination requires more sampling, testing, and monitoring. Agricultural regions, chemical manufacturing sites, and military establishments are all areas of concern, especially if drinking water is drawn from groundwater.