Amorphous Mineral Colloids of Soils of the Pacific Region and Adjacent Areas' YOSHINORI KANEHIR02 and LYNN D. WHITIIG 3 THE PRESENCE of amorphous mineral colloids have pointed out that the amorphous consti ­ in soils and geologic formations is not as un­ tuents make up a sizeable fraction in many soils common as was first believed in the early years occurring in Hawaii, Japan, New Zealand, Ore­ following the acceptance of the clay mineral gon, and other Pacific areas. These amorphous concept. In the early reports the occurrence of mineral colloids play a prominent role in soil amorphous material was associated with only a formation and also impart certain distinctive few rare and isolated clay materials. Because and unique properties to the soil. Thus, a re­ amorphous colloids are not the major component view of this nature appears justified. in most soils and their presence may be found in relatively low concentrations, if found at all, NOMENCLATURE OF AMORPHOUS COLLOIDS their detection has been difficult. Moreover, whereas crystalline clay minerals are relatively The isolation and description of amorphous -uniform in composition, the amorphous mate­ colloids have been difficult because of the great rials exhibit a varying degree of composition variability in materials. Moreover; early sam­ and poor degree of crystallinity, further adding ples classified as "amorphous" were actually to the difficulty in their identification. Often found to be finely crystalline with modern X-ray their presence has been suggested only because diffraction methods. Stromeyer and Hausmann mineral allocations of crystalline materials failed first used the name allophane to describe amor­ to add up to 100 per cent. In recent years im­ phous material lining cavities in marl in 1816. provement in the use of techniques such as Since that time many related materials have X-ray diffraction, infrared absorption, electron been called allophane and this term has become microscopy, and surface area determination, has associated with amorphous constituents of clay. made it possible to make significant progress in Ross and Kerr (1934) described allophane as the study of amorphous colloids. essentially an amorphous solid solution of silica, Much of the research dealing with amorphous alumina, and water having no definite atomic mineral colloids in soils has been conducted by structure, and they applied the term allophane soil scientists working in the Pacific region or to a great number of amorphous clay materials in its adjacent areas. The leadership in this field regardless of their composition. They studied definitely belongs to this group of researchers. five specimens of allophane, all essentially hy­ It is the object of this paper to review and drous aluminum silicates, and found that Si0 discuss the contributions of these workers in 2 ranged from 25 to 34 per cent, Ah03 from 30 order to obtain a better perspective of this very to 36 per cent, and H from 31 to 38 per important fraction of soils. These investigators 20 cent. The New Zealand workers (Fieldes et al., 1952, 1954; Birrell and Gradwell, 1956) have 1 This paper is based on part of a joint report by _members of the Clay Mineralogy Work Group of the used the term amorphous colloidal hydrous ox­ Western Soil and Water Research Committee. Manu­ ides apart from the term allophane in their script received October 3, 1960. description of amorphous clays. With allophane, 2 Department of Agronomy and Soil Science, Uni­ which is considered to be one of the most im­ versity of Hawaii, Honolulu. 3 Department of Soils and Plant Nutrition, Univer­ portant amorphous minerals, Fieldes (1955, sity of California, Davis, California. 1956) has preferred to recognize three distinct 477 478 PACIFIC SCIENCE, Vol. XV, July 1961 forms : allophane A, allophane B, and the in­ two hydrol humic Iatosols from the island of termediate form, allophane AB. In classifying Hawaii. The authors noted that the allophane the clay minerals Grim ( 1953) has included found in the subsoil of one of these soils was only the allophane group under the amorphous very similar to allophane from Woolwich, Eng­ clay minerals. Brown (1955) in his proposed land (Kerr, 1951 ) . Gibbsite and goethite were nomenclature has divided the amorphous min ­ reported to make up the bulk of the remaining erals into oxides, silicates, and phosphates. In clay fraction. They also investigated the low this system, allophane is included in the silicates. humic latosols and reported that the dominant minerals are of the kaolin family. Up to 10 per cent allophane was found to occur in the OCCURRENCE OF AMORPHOUS COLLOIDS clay fraction of this group of soils. Kelley and Page (1943) in their mineralog­ In a subsequent paper (1955) the same ical investigation encountered two soils from authors reported on a humic ferruginous late­ Naalehu and South Point on the island of Ha­ sol from the island of Maui which showed that waii that exhibited very high cation exchange almost 30 per cent of the clay fraction in the capacities, 120 m.e. and 88 m.e. per 100 g., subsoil was composed of allophane. respectively. They reported that differential The occurrence of allophane in some soils of thermal analysis showed pronounced endother­ northwestern Oregon was suggested by Whittig mic peaks at 160 0 C. for these two soils in et at. (1957). These soils, members of the addition to showing weak X-ray diffraction pat­ Cascade and Powell series, contained relatively terns. These inxestigators, therefore, concluded high percentages of alkali-soluble silica and that the high cation exchange properties were alumina. The amorphous alumino-silicate, in related to the presence of considerable amor­ these soils was formed by weathering of aeolian. ' phous material. Included in this study were volcanic ash. soils from Vale, Oregon, and the Mojave Desert, In earlier work, Whittig (1954 ) reported which also gave very indistinct X-ray lines and the occurrence of a more stable form of allo­ showed low temperature breaks , inferring the phanein two 'humic ferruginous latosols of Ha­ presence of amorphous material. waii. The allophane of ' these soils had a rela­ Dean ( 1947) in his D.T.A. study of a num­ tively low cation exchange capacity (of the ber of Hawaiian soils derived from ash and order of 10 m.e. per 100 g.) and resisted solu­ lava found that many of these soils contained ' tion in boiling Na2C03 solution. almost no crystalline clay minerals. In addition More recently Bates (1961) described the some showed almost no hydrous oxides. It was presence of mineral gels in Hawaiian soils which previously shown by Ayres (1943) that some are mixtures of aluminum, iron, silica, and ti­ of these same soils possess very high inorganic tanium compounds. The gel material is very cation exchange capacities. Dean concluded that reactive chemically and gives rise to inorganic it was possible that some of these soils contain and organomineral complexes in the colloid alterations of the kaolin minerals. fraction. Tanada (1950) divided Hawaiian soils into Matsusaka and Sherman (1960) have re­ five groups on the basis of chemical analyses ported that the iron hydroxide and oxide of the and dehydration studies . He obtained similar amorphous mineral colloid fraction of Hawai­ high cation exchange capacity values for the ian lateritic soils will form strongly magnetic two soils, Naalehu and South Point, that Kelley iron oxides on dehydration. This may help ex­ and Page (1943) had previously reported. How­ plain the magnetic properties of weathered ever, Tanada did not draw any conclusions re­ ferruginous geological formations. garding the cause of such high values. In Japan Sudo ( 1954) , Sudo and Ossaka Tamura, Jackson, and Sherman ( 1953) em­ ( 1952), and Aomine and Yoshinaga (1955) ployed X-ray, chemical, thermal, and infrared have pointed to allophane as the dominant con­ techniques and found up to 30 per cent allo­ stituent of Ando soils which are formed from phane in the less than 0.2 micron fraction of volcanic ash. These soils are characterized by a Amorphous Mineral Colloids-KANEHIRO and WHITIIG low bulk density, a high organic carbon con­ ( 1951 ) offered confirmatory evidence by ab­ tent, and low base saturation. These properties sorption spectra that allophane is not a mixture are attributed to the preponderance of allo­ of alumina and silica. phane. The Ando soils and related types are Tamura et at. ( 1953) assigned allophane to found associated with the Pacific ring of vol­ weathering stage 11 or the gibbsite stage in the canic activity. These Japanese workers have weathering sequence of clay-size minerals as found that the fine clay fraction of the Ando presented by Jackson et at. (1948 ). They noted soils is characterized by being amorphous to that the trend for increased gibbsite with in­ X-rays and possesses medium-to-high cation ex­ creased rainfall is very marked in passing from change capacities and high phosphate- and the low humic latosols to the hydrol humic ethylene glycol-retention values. latosols. With this increase in gibbsite is an The New Zealanders have also worked ex­ associated increase in allophane. tensively on the identification of amorphous con­ A mechanism for the transition of alumina stituents. In 1952 Birrell and Fieldes (1952) and silica through allophane to kaolin was pro­ and Birrell (1952) identified the presence of posed by Tamura and Jackson (1953). The amorphous material, principally allophane, in steps are as follows: (1) amorphous hydrous soils derived from rhyolitic and andesitic ash. alumina crystallizes to a gibbsite structure; (2) The allophane was found to be present mainly with partial dehydration, hydroxyls in the gibb­ in the clay fraction although it was inferred site octahedra are replaced by oxygens of the that it was present to some extent in the silt silica tetrahedra; (3 ) this process occurs in the fraction.