A Convention for Expressing the Mass of High Molecular Weight Compounds

Letter to the Editors

The difficulties we have all experienced in handling compounds of high molecular weight are steadily being overcome and there are now appearing in the literature, particularly in Biomedical Mass Spectrometry, reports of mass spectra covering more than 500 mass units.

It is becoming increasingly clear that there is little agreement amongst workers in the field as to how the mass possessed by all atoms except carbon can be handled.If one calculates the formula mass of the species C15H26N2O, whose nominal mass is 250, this can be accomplished in two distinct ways: (1) The fractional masses of the appropriate atoms can be rounded (usually down) to the nearest integer and the formula mass can then be calculated using H = 1, C = 12 and so on. (2) Alternatively, the accurate masses of the atoms can be used. This will lead to a formula mass of 250.2045 which is then rounded to the nearest whole number. Either method leads to a nominal mass of 250 for the species in question.

At masses above 436, however, these two methods can lead to different answers, most commonly due to the high mass defect of hydrogen. Thus the precise mass of C31H64 is 436.5008, and so its nominal mass as calculated by method (2) above would be 437, which is at best, a misleading result. Examples of this difficulty are beginning to appear in the literature and we might cite the paper of G. W. Wood and P.-Y. Lau Biomed. Mass Spectrom. 1974, 1, 154. This paper deals with, inter alia, the ion of atomic composition C40H80NO8P. Use of method (1) described above leads to a mass of 733 for this formula, but the authors use method 2 and so arrive at a mass of 733.6 which they subsequently round off to the nearest integer. arriving in this case at a value of 734, which might erroneously suggest that the species contains either zero or an even number of nitrogen atoms.

In a second example, the precise mass of the heptaiodotolyl alkane, C39H65I7 is 1421.8400. Loss from such a molecular ion of C31H63 (435.4930) would produce the benzyl daughter ion C8H2I7 (986.3470). If these numbers are now rounded (as required by method 2 above), one arrives at the result 1422 - 435 = 986, which is arithmetically absurd, as well as being chemically misleading. Such errors will be common if the rounding of large fractions is carried out while plotting a mass spectrum as a bar graph. This is admittedly a contrived example, but when such large cumulative mass defects are involved, confusion as to the nominal mass of either the daughter ion or the neutral fragment that is lost will be likely.

Neither of these methods is fully defensible, and since they can lead to different answers, we suggest that a convention be adopted by both authors and editors of this journal. The convention is simply this: For the sake of discussion in papers, the nominal mass of any ion shall be defined as that integral mass that is calculated using integral atomic masses. Thus C31H64 shall be regarded as having a nominal mass of 436; C124H250, whose accurate mass is 1739.9563 shall have a nominal mass of 1738 and the species C40H80NO8P will have a nominal mass of 733 and a precise mass of 733.5622.

Adoption of this convention will lead to the following advantages: (1) No decimal places will be necessary in specifying a mass and there will be no ambiguity as to the mass specified for ions and for most small fragments lost from parent ions. (2) The 'nitrogen rule' will never be invalidated. (3) The plotting of mass spectra in the familiar bar graph form will be straight forward.

On the other hand, if an accurate mass measurement has been made, this should be quoted in full as the mass of the ion in question. If the mass of an ion has been determined as, say, 750.7 (by, for example, a mass marker) then its mass should be so quoted unless it is known that the decimal part represents a positive mass defect (as in, for example, a hydrocarbon). In such a case, the number can be adjusted down. It might be noted in passing that the decision as to the sign, if not the magnitude, of the mass defect is almost always straightforward in organic mass spectrometry. Ions with more than six iodine atoms are the only exceptions that come readily to mind.

This is a surprisingly complicated problem and we are not certain that the best solution to it is the one proposed here. We would, therefore, urge any of our colleagues who are interested in the question to communicate their thoughts on the subject so that a consensus solution may be reached.

Yours truly,

H. M. Fales
S. R. Heller
G. W. A. Milne
T. Sun

Department of Health, Education and Welfare
Public Health Service
National Institutes of Health
National Heart and Lung InstituteBethesda
Maryland 20014
15th June 1974