Ionized nitrogen mono-hydride bands are identified in the presolar and carbonado diamond spectra


  • Jozsef GARAI

    1. Department of Mechanical and Materials Engineering, Florida International University, Miami, Florida 33199, USA
    2. Department of Mechanical and Structural Engineering, Ybl Miklós Faculty, Szent István University, Budapest, Hungary
    Search for more papers by this author



Abstract– None of the well-established nitrogen-related IR absorption bands, common in synthetic and terrestrial diamonds, have been identified in the presolar diamond spectra. In the carbonado diamond spectra, only the single nitrogen impurity (C center) is identified and the assignments of the rest of the nitrogen-related bands are still debated. It is speculated that the unidentified bands in the nitrogen absorption region are not induced by nitrogen, but rather by nitrogen-hydrides because in the interstellar environment, nitrogen reacts with hydrogen and forms NH+; NH; NH2; NH3. Among these hydrides, the electronic configuration of NH+ is the closest to carbon. Thus, this ionized nitrogen-mono-hydride is the best candidate to substitute carbon in the diamond structure. The bands of the substitutional NH+ defect are deduced by redshifting the irradiation-induced N+ bands due to the mass of the additional hydrogen. The six bands of the NH+ defects are identified in both the presolar and the carbonado diamond spectra. The new assignments identify all of the nitrogen-related bands in the spectra, indicating that presolar and carbonado diamonds contain only single nitrogen impurities.


Nitrogen is the most important impurity in diamonds. The content and the type of the nitrogen defects are even used to classify the diamonds (Davis 1977). If the nitrogen content is significant, typically measured in ppm but can be as high as 1%, then the diamonds are type I while diamonds with undetectable or very small amounts of nitrogen are type II. When a diamond forms, nitrogen atoms can replace the carbon atoms, forming an isolated single impurity. Thus, newly formed diamonds, like synthetic diamonds, almost exclusively contain single nitrogen impurities. Energetically, a higher level of aggregation of nitrogen impurities is favored, single nitrogen impurities tend to unify and form higher levels of aggregation. The bonds between the carbon atoms in diamonds are very strong resulting in very slow diffusion, even on the time scale of a billion years. Elevated pressures and temperatures increase diffusion rates resulting in higher levels of aggregation, clustering two, four, or more nitrogen atoms. The aggregation level of nitrogen can be used to estimate the mantle residence time for terrestrial diamonds. The lack of higher level nitrogen aggregation indicates that diamonds were not affected by high pressures and temperatures and long annealing times after formation. The nano-size of the presolar diamonds, which typically contain in the order of 103 C atoms and 10 N atoms (Jones and d’Hendecourt 2000; Jones 2001), might put constraints on the nitrogen aggregation.

About 98% of natural terrestrial diamonds contain defects with two and four aggregated nitrogen atoms. Based on the aggregation level of the nitrogen, type I diamonds are subdivided into two subgroups. Type Ia diamonds (98%) contain clustered nitrogen atoms, and type Ib (0.1%) contain scattered or single nitrogen atoms. Type II diamonds are also subdivided into type IIa (1–2%), which is almost pure carbon, and type IIb (0.1%) containing boron atoms. The percentages refer to the natural occurrence of different diamond types. The aggregation level of nitrogen can be identified by infrared absorption. The single (C center), pair (A center), four (B center) nitrogen and higher aggregations, like platelets each induce well-defined characteristic infrared absorption bands. These nitrogen-related bands are present in terrestrial and synthetic diamonds, but lacking in the presolar diamond IR spectra (Lewis et al. 1989; Colangeli et al. 1994; Koike et al. 1995; Mutschke et al. 1995, 1996; Hill et al. 1997; Andersen et al. 1998; Braatz et al. 2000).

The spectra of carbonado diamond contain the bands relating to substitutional N defects at 1130 cm−1 and 1344 cm−1; however, the spectra are very different from type Ib diamonds (Fig. 1) and contain many additional unidentified bands in the nitrogen region. No other bands relating to higher nitrogen aggregation are present in the spectra. The rest of the characteristic bands identified in the carbonado diamond spectra (Garai et al. 2006; Kagi and Fukura 2008) are similar to the presolar bands; therefore, these diamonds are also included in this investigation. It should be noted that the origin of carbonado remains enigmatic. The almost identical IR spectra are not an ultimate proof for the same origin and formation. The assignments of the nitrogen-related bands in the presolar and carbonado diamonds spectra are still debated and the exact configuration of N is not known. The nitrogen in the presence of hydrogen can react with hydrogen forming nitrogen hydrides (Mitchell et al. 1978; Huebner and Giguere 1980; Adams et al. 1984). The exact mechanisms for formation of nitrogen hydrides in space environment are still debated; however, it is widely agreed and supported by observations that a significant part of the nitrogen in space is in the form of nitrogen hydrides (e.g., Swings et al. 1941; Cheung et al. 1968; Schmitt 1969; Meyer and Roth 1991; Crawford and Williams 1997; Meier et al. 1998; Hily-Blant et al. 2010; Persson et al. 2010). It is speculated that if nitrogen is hydrogenated, then nitrogen hydrides could substitute carbon when diamond is formed. This possibility is investigated in detail.

Figure 1.

 Midinfrared absorption spectra of carbonado diamond from the Central African Republic (CAR) in the C-N vibrational region (solid line) are plotted. The experimental procedure is described in Garai et al. (2006). Among the characteristic terrestrial diamond absorption bands, only the substitutional N0 bands (C Center) are present in the spectra. The assignments of the rest of the bands are still debated. For comparison, the characteristic absorption spectra of Ib diamond containing substitutional N (C Center) defects (dashed line) (Field 1992) and the spectra of Allende DM1 microdiamond (dotted line) (Lewis et al. 1989) are also shown. The image shows a carbonado from the Bangui region, Central African Republic. The patina of the surface is a characteristic feature of carbonado diamonds.


To extract the presolar diamonds from meteorites, the samples go through severe chemical treatments, which alter the original spectra (Lewis et al. 1987; Anders and Zinner 1993; Amari et al. 1994). Based on these contaminations, it has been concluded that in the presolar diamond IR spectra, only the absorption features relating to nitrogen atoms (Braatz et al. 1999) and to the multiple phonon absorption of diamonds can be considered as intrinsic properties of the diamond. In this study, the absorption features of nitrogen impurities are investigated.

The complete data set of previous presolar and carbonado diamond IR spectra reported in the literature was collected. The characteristic bands are listed in Table 1.

Table 1.   Reported spectral bands for presolar and carbonado diamonds in cm−1 and their proposed assignments.
AllendeMurchisonOrgueilCarbonadoTerrestrial (assignments)Assignments to presolar and carbonado diamonds by
  1. aThe numbers represent the averages of the bands from the three spectra. The numbers in parentheses relate to bands observed in two of the spectra.

1Lewis et al. 19892Koike et al. 1995;3Andersen et al. 1998;4Braatz et al. 2000;5Colangeli et al. 1994;6Mutschke et al. 1995;4Braatz et al. 2000;7Hill et al. 1997;8Garai et al. 2006aInt. %9Kagi and Fukura 2008;10McNamara and ref. therein 200311Lawson et al. 1998;Previous studiesThis paper
321031153236 3164   (∼3200)   1, 2, 3, 5, 6, 7N-H StretchN-H Stretch
163216121616 1623 1650261640 1, 5Aromatic C=C/C=O stretch, 2C=O/N-H stretching, 3O-H bend (in H2O)N-H Bending
     14654  3C-H deformation (CH3/CH2)Subst. NH+; C-H def. (CH3/CH2)
 1399   1420141438 1, 3, 5Interstitial N/C-H deformation (CH3)Subst. NH+; C-H def. (CH3)
        (1332)313341344 (Subst. N)11Subst. NH+
       1289  12841282 (A aggregates)102, 5, 6, 7C-O/C-N Stretch 3C-O/C-N/C-C Stretch/Chem. treat.7/A cent.
1234  1210  1220 122011 1220 (C-N Stretch)10 (C-N0) Stretch
11731178 1150 11751155 (1163)40  Interstitial N1 C-O Stretch/Interstitial N(C-NH+) Stretch
 11221122    1125(1132)3411261130 (Subst. N)10Subst. N0
110811081109     11001001100 4C-O StretchSubst. NH+
10901080109010841084109010881072    4C-O(H), CF2Chem. treatment4
1028 1054     103045   Subst. NH+
        90972   Subst. NH+

In the nitrogen region, the band at 1085 cm−1 has been proposed that results from chemical treatment (Braatz et al. 2000). The spectra of carbonado diamond contain all of the reported bands of presolar diamonds with the exception of the 1085 cm−1. As the carbonado samples were not chemically treated (Garai et al. 2006), the lack of the 1085 cm−1 band is consistent with the proposed chemical treatment-induced origin of this band.

The band at 1285 cm−1, characteristic of A center in terrestrial diamonds, is reported in the Orgueil and the carbonado diamond spectra. This band is greatly diminished by further oxidation of the nano-diamond residues of Orgueil chondrite, indicating that the band is an artifact resulting from the chemical treatments (Hill et al. 1997). This band is present in all of the chemically treated carbonado diamond spectra (Kagi and Fukura 2008); however, it is also detected in one of the nontreated carbonado spectra. Thus, the possibility that this band might be induced by A centers cannot be excluded.

The characteristic bands of C center at 1130 cm−1 and 1344 cm−1 are present in the spectra of carbonado diamond. In the Allende and Orgueil spectra, there is a band at 1122–1125, but the 1344 cm−1 band is missing. Thus, the presence of the substitutional nitrogen impurities (C center) is debatable in presolar diamonds.

None of the nitrogen-related bands, common in terrestrial and synthetic diamonds, are uniquely identifiable in any of the presolar diamond spectra. Based on the lack of the well-established nitrogen-related bands, it is speculated that the unidentified bands in the nitrogen absorption region might not be induced by nitrogen, but rather by nitrogen-hydrides, which are common in the interstellar environment.

The bonds between N and C are covalent. Covalent bonds are directional and dictate the spatial arrangements of the atoms. Carbon atoms in the diamond structure have tetrahedron coordination based on the spatial distribution of the sp3 hybridized electrons. To fit into the diamond lattice, the spatial distribution of the electrons of nitrogen has to be the same. The electronic structure of Nitrogen is 2s2 2p3. The three bonding domains of the p electrons and the one nonbonding domain of the loan electron par electrons form a tetrahedron skeleton structure, which fits into the diamond lattice. Thus, the single nitrogen impurity in the diamond lattice is bonded to three of the neighboring carbon atoms. This conclusion is supported by the existence of the NV defects. If nitrogen were bonded to four carbon atoms, then the NV defects in diamond would not be stable. The tetrahedron skeleton structure of the diamond lattice can also be achieved through the ionization of the nitrogen. If nitrogen is ionized, then the remaining four electrons are hybridized to sp3. Thus, N+ fits comfortably into the diamond lattice by forming four bonds. It can be predicted that among the available nitrogen hydrides (NH+; NH; NH2; NH3), the coordination of NH+ would preferentially fit into the bulk diamond lattice. Thus, it can be predicted that NH+ could substitute carbon in the diamond lattice and form an NH+ defect when nitrogen hydrides are present at the formation of the diamonds.

The absorption features of the NH+ defects are not known. In an attempt to identify the features NH+ defects, the bands of N+ defects were redshifted by increasing the mass of nitrogen with one atomic unit due to the mass of the additional hydrogen.

The ionization of the substitutional N (C center) and conversion into substitutional N+ (C+ center) has been predicted (Lawson and Kanda 1993), and later experimentally verified by the irradiation of type Ib diamonds. The irradiation-induced C+ center exhibit a sharp peak at 1332 cm−1, and broader and weaker peaks at 1115 cm−1, 1046 cm−1, and 950 cm−1 in their spectra (Lawson et al. 1998). The same bands were observed in the spectra of synthetic diamonds after radiation damage by electrons (Collins et al. 1988) along with two additional bands at 1450 cm−1 and 1502 cm−1, which have been assigned to C-N. All the bands of the C+ center are redshifted in the carbonado-diamond spectra with the exception of the band at 1332 cm−1 (Fig. 2a). It has been shown experimentally that N15 isotope doping does not have an effect on the radiation-induced band at 1332 cm−1 (Samoilovitch et al. 1975). Thus, no redshift is expected for the 1332 cm−1 band from the additional mass of hydrogen. Using a simple force constant model, the rest of the irradiation-induced N+ bands are redshifted by adding one atomic unit to the mass of nitrogen, due to the mass of the additional hydrogen. The calculated absorption bands of the NH+ defects are listed in Table 2. The employed simple force constant model can be used, and gives reasonably good results, when the bonds are uniform and spherically symmetric. The permanently positively charged nitrogen with tetravalent bonds satisfies these requirements.

Figure 2.

 Midinfrared absorption spectra collected from thin section of Brazilian (BR) and Central African Republic (CAR) carbonado diamond are plotted (Garai et al. 2006). The spectra are an intrinsic property of the diamond because no chemical treatment was used in the preparation. a) The irradiation-induced N+-related bands are present, but redshifted in the carbonado spectra. The observed redshift is consistent with one atomic mass unit increase in the nitrogen, which can be caused by the hydrogenation of the ionized nitrogen. b) The presence of hydrogenated nitrogen is supported by the N-H stretch and bending bands at around 3200 cm−1 (3.12 μm) and 1652 cm−1 (6.05 μm), respectively. The assigned vibrational regions are identified.

Table 2.   The irradiation-induced N+ bands are redshifted by adding one atomic unit to the mass of nitrogen, due to the mass of the additional hydrogen. The redshifts were calculated using simple force constant model.
AssignmentFrequency (cm−1)
C-N+ (μm)C-NH+ (μm)
Irradiation-induced substitutional N+(C+-center)950 (10.53)935 (10.70)
1046 (9.56)1030 (9.71)
1115 (8.97)1098 (9.11)
1450 (6.90)1428 (7.00)
1502 (6.66)1479 (6.76)

The agreement between the redshifted substitutional N+ bands and the bands of the presolar and carbonado diamond spectra is good. Based on this good agreement, it is suggested that the ionized nitrogen-mono-hydride complex bonded into the four neighboring carbon atoms, and that the bands observed in the nitrogen absorption region (Table 1) can be assigned to substitutional NH+ defects.

The presence of NH+ defects in carbonado diamonds is consistent with hydrogen isotopic composition investigations, which report about 70 ± 30 ppm of hydrogen, relating to the bulk of the diamond (Demeny et al. 2011).

The presence of NH+ defects in presolar and carbonado diamonds is also supported by the observed N-H-related bands. Both presolar (e.g., Andersen et al. 1998) and carbonado diamond spectra (Fig. 2b) contain intense N-H-related stretching vibration bands in the region of 3100–3400 cm−1 and an N-H bending band at 1650 cm−1. These bands are absent or significantly weaker in the spectra of terrestrial and CVD diamonds (e.g., McNamara et al. 1994).

The two-step formation of NH+ defect, ionizing the substitutional N0 or C center by irradiation and then hydrogenating the ionized nitrogen, might also be possible. However, no N+-related bands can be identified in any of the spectra. The complete lack of this “middle stage” defects in this two-step formation process indicates that the NH+ defects were most likely formed by a one step process. Thus, the ionized nitrogen mono-hydrides were substituted into the diamond structure at the time of the formation.

The NH+-related bands at 1420 cm−1 and 1465 cm−1 overlap with C-H deformation bands; therefore, both NH+ and C-H deformations are assigned to these bands.

Despite the similarities between the presolar and carbonado diamonds’ IR spectra, there are subtle differences. The substitutional NH+ band at 1332 cm−1 is reported only in the carbonado spectra. The nitrogen in carbonado diamond is dominantly present in the form of NH+ defects, but substitutional N0-related bands are also detectable in the spectra, indicating that both nitrogen-hydrides and nitrogen were available at the time of the diamond formation. The ratio between N0 and NH+ defects in the diamond could be used to set constraints on the environment in which the diamonds were formed.


It is suggested that the unidentified bands observed in the nitrogen region of presolar and carbonado diamond IR spectra arise from substitutional NH+ defects. In the carbonado spectra, the N0 bands are also detectable, which indicate that both nitrogen and ionized nitrogen mono-hydride were present at the formation of the diamond.

The bands of the substitutional NH+ defect are deduced by redshifting the irradiation-induced N+ bands due to the mass of the additional hydrogen. The six bands assigned to the NH+ defects are identified in both the presolar and the carbonado diamond spectra. With the new assignments, all of the nitrogen-related bands in the spectra are identified, and it is shown that presolar and carbonado diamonds contain almost exclusively single nitrogen impurities; therefore, they can be classified as type Ib.

Acknowledgments— I thank Andriy Durigin and Subrahmanyam Venkata Garimella for the inspiring discussions and for their thoughtful comments. I also thank Mike Sukop and Istvan Kovacs for reading and commenting on the manuscript. This manuscript benefited from the careful review of Hugh Hill and Anthony P. Jones.

Editorial Handling— Dr. Scott Sandford