Dialuminium nickel tetraoxide

Adsorption / desorption

Description of key information

Partition coefficients for suspended matter, soil, STP, sediments in freshwater and in coastal waters are available. Log Kp values for all types ranged from 2.86 to 4.42.

Additional information

No adsorption/desorption data are available on the substance yet various reliable data exist for nickel (measured as environmental concentrations) and different analogue nickel substances showing statistical or conservative partition coefficients for suspended matter, soil, STP, sediments in freshwater and in coastal waters. Log Kp values for all types ranged from 2.86 to 4.42.

Kp- freshwater suspended particulate matter (SPM)

Reported Kp-values between water and suspended particulate matter range from 692 to 309,030 L/kg (Log Kd: 2.84-5.49). The data are extracted from studies reporting water/suspended matter partitioning coefficients for the Netherlands, UK, France, Czech Republic, and Italy.

 

Reported log Kd, suspended particulate matter (SPM) values for Ni in freshwater surface waters in Europe 

 

The partitioning coefficients between water and SPM were fitted to a cumulative distribution function (CDF) (Heijerick and P. Van Sprang, 2004; European Nickel Risk Assessment). The 10^th, 50^th, and 90th percentiles of the CDF were selected as final suspended solids partitioning coefficients for the local/regional exposure analysis:

 

Kpsusp = 26,303 L/kg (log Kp susp= 4.42 L/kg) (50^th percentile)

Kpsusp,min = 5,754 (log Kpsusp min = 3.76 L/kg) (10^th percentile)

Kpsusp,max = 117,490 (log Kp susp max = 5.07 L/kg) (90^th percentile)

 

The 50^th percentile value of the distribution function represents a typical suspended matter partition coefficient for EU waters and will be used for the derivation of local and typical regional.

 

Kp-freshwater sediment

Reported Kp-values between water and sediment are presented in the first table below (data extracted from studies reporting water/sediment partitioning coefficients for the Netherlands and UK) and in the second table below (river sediment data compiled by Gunn et al., 1992). The Kp values reported by Gunn et al. (1992) represent the individual or mean value of British-German rivers. Individual data that were used for the determination of mean data were not reported. The observed values range from 770 to 24,300 L/kg (Log Kp: 2.89-4.39).

 

reported log Kp-values on sediment

Water body, location	log Kp l/kg	Remarks	Reference
Mersey river,	3.71	Modelled value	Turner et al.2002
Dutch freshwater	3.72	Modelled value	Stortelder et al. 1989; in Crommentuyn et al. 1997
Dutch freshwaters: Nieuwe Merwede Rijn - Hagestijn Nieuwe Waterweg Oude Maas Waal Rijn – Lobith Ketelmeer Maas- Eijsden Ijsselmeer Haringvliet	3.79 3.86 3.88 3.88 3.89 3.92 3.96 4.04 4.06 4.34	Mean value Mean value Mean value Mean value Mean value Mean value Mean value Mean value Mean value Mean value	Van Der Kooij et al.1991
Range	3.71-4.34

 

 

Reported Kp values for nickel in river sediments (Gunn et al. 1992)

Site – river / estuarine sediments	Number of measurements	Mean Kp value l/kg	Log Kp sed	Reference
Trent (Althorpe)	3	2750	3.44	Comber et al. 1995
Mersey(Howley weir)	1	7690	3.89	Comber et al. 1995
Rhine (Lauterbourg, Fr/Ger border)	8	24300	4.39	Kern et al.1998*
Ythan (Scotland)	1	770	2.89	Comber et al. 1995
Ouse (Selby)	1	1000	3.0	Comber et al. 1995
Mean		7300	3.9

 

The partitioning coefficients between water and sediment were fitted to a cumulative distribution function (CDF) (Heijerick and P. Van Sprang, 2004; European Nickel Risk Assessment). The 10^th, 50^th, and 90th percentiles of the CDF were selected as final sediment partitioning coefficients for the local/regional exposure analysis:

 

Kpsed = 7,079 L/kg (log Kpsed = 3.85) (50^th percentile)

Kpsed, min = 2,138 L/kg (log Kpsed min = 3.33) (10^th percentile)

Kpsed, max = 16,982 L/kg (log Kpsed max = 4.23) (90^th percentile)

 

It should be noted that the distribution in the above tables is based on data that, with the exception of one value, only represent surface waters located in The Netherlands and the United Kingdom. The representativity of this distribution for the general case is evaluated using the distribution functions of ambient Ni-concentrations in water and sediments that were generated using extensive EU-monitoring data sets (European Union Nickel Risk Assessment, 2008/2009). Based on the 5^th/95^th, 50^th/50^th, and 95^th/5^th percentiles of the distributions of Ni in water/sediment, respectively, a range of theoretical Kp-values was defined, with a median log Kp of 4.07 and an estimated 10/90th percentile range of 3.62-4.56. These percentiles are in line with the values given and indicate that the natural variability of Kp sediment is well covered in this distribution.

Kp-Marine (STP and Sediment)

Marine partitioning coefficients for sediment and SPM are based on the study by Stuer-Lauridsen et al. (1996). The following table represents the data used to derive the marine partitioning coefficients:

Partitioning coefficients Kp in Danish brackish coastal water:

Location (1995)	Sediment	Suspended matter	Total water	Filtrate	Porewater	Sediment/total water	SPM/filtrate	Sediment/ porewater
	Mg Ni/Kg dw	Mg Ni/kg dw	ug Ni/L	ug Ni/L	ug Ni/L	Kd, l/kg	Kd, l/kg	Kd, l/kg
Roskildefjord l	16.40	2.97	0.52	0.79	26.47	15710	3070	620
Roskildefjord ll	15.80	5.60	0.66	0.45	2.04	23910	8840	7680
Øresund	18.07	7.40	2.56	1.06	0.53	7070	6960	34260
Mean						15563        (log 4.2)	6290          (log 3.8)	14190             (log 4.2)

 

In Danish coastal waters (Table above) three locations Kpsusp ranged 3070 to 8840 l/kg with a mean value 6290, i.e. logKpsusp.marin 3.8.

The range from three locations had Kd values from 7070 to 23910 l/kg with the average 15560 (Table above), i.e. logKpsed.marin 4.2.

Kp- aquatic to STP:

Partitioning coefficients for aquatic to STP are based on the studies by Moriyama et al. (1998), Coumber and Gunn (1992, 1994) and Gould and Genetelli (1992). The table below represents the data used to derive the aquatic to STP partitioning coefficient of LogKp = 3.4

Partition coefficients Kp for STP:

STP	Coefficient	Range, l/kg	Value, l/kg	Log value	Reference
Kprs	solids-water in raw sewage sludge (m^3/m^3)	1720 – 2569	2333 ***	3.4	Moriyama et al.1989, Comber and Gunn 1994
Kpa	solids-water in activated sewage sludge (m^3/m^3)	126 – 1259 453 – 1935	1000	3	Gould and Genetelli 1982 Comber and Gunn 1992

*** No published data available on Kp for raw sewage – data obtained from combination of total Ni in raw sewage and Ni in faeces, assuming 100 mg/L SS

Nickel removal rates used for determining the STP removal parameter were limited to those reported after 2000 because removal rates increased with time for Denmark and the Netherlands. Ni removal rates in sewage treatment plants from recent years (2000-2002) are situated between 41% (DEPA, 2002) and 50% (CBS, 2004). Based on the available data, the value 40% removal represents a reasonable worst case removal of Ni in STP in the EU.

Kp- Soil:

The variation observed in the American log Kp soil values may relate to the variation in pH values. The log Kp average was 2.08 in 11 soils with a pH range 4.3 to 4.85 (Buchter et al., 1989) whereas the log Kp average was 0.77 in 15 soil with a pH range 4.2 to 6.5 (King 1988). Anderson et al. (1988) reported log Kp values in spiked Danish agricultural soils varying between 1.0 and 3.0. As the equilibration times in the Buchter et al. (1989) and King (1988) studies were relatively short, the studies performed by Janssen et al. (1997), De Groot et al. (1998) and Sauvé et al. (2000), describing field-derived partitioning coefficients, are more relevant to the risk assessment. Janssen et al. (1997) study on 20 contaminated Dutch soils found a median value log Kp 3.17. This value is close to the log Kp value of 2.86 (Aqua regia digestion) found by De Groot et al. (1998). Sauvé et al. (2000) reported a median log Kp value of 3.37 for a wide range of soil types. However, since it is clear that the Kd levels derived in this study are high compared to other studies and it is unclear what type of extraction techniques exactly are used, it is assumed that stronger digestion methods such as HF extraction or X-ray fluorescence -determining the metal fraction built into the crystal structure of the soil minerals, and hence not relevant for RA purposes- may be incorporated in these results. As the De Groot et al. (1998) study covered 46 European soils, although mainly Dutch soils, and aqua regia extraction was used as digestion method, this value is preferred for the risk assessment. The partition coefficient determined by aqua regia extraction is selected since for environmental purposes, the absolutely maximum fraction that may be released in time-determined through aqua regia digestion- is of interest (FOREGS Geochemical Baseline Programme–Analytical Manual, 2001). This digestion method is harmonised as an International Standard (ISO 11466) and is applied in most EU-countries. This extraction method is also applied in the ongoing scientific research program on Ni toxicity and bioavailability in soils. The log Kpsoil of 2.86 (Aqua regia extraction) is used as input value for the exposure modelling,

i.e.log Kpsoil 2.86