Material guide: benzyl benzoate, classified by the database as both a fragrance material and a carrier solvent

In short: one of the few materials I have written up whose category field itself says carrier solvent, while also having an odour of its own. What to know before buying is the melting point and the specific gravity.

Fields

Field	Value
CAS	120-51-4
First synonym	phenylmethyl benzoate
Formula	C14H12O2, MW 212.248
Database category	flavor and fragrance agents, carrier solvents
Appearance	colorless to pale yellow clear oily liquid to solid (est)
Assay	99.00 to 100.00
Melting point	18-21 C
Boiling point	323-324 C @ 760 mmHg
Flash point	158.00 C
logP	4.00 (est)
Vapour pressure	0.000250 mmHg @ 25 C
Vapour density	7.3 (air = 1)
Specific gravity	1.113-1.121 @ 25 C
Refractive index	1.566-1.571 @ 20 C
Odour type	balsamic
Odour	sweet; balsamic; oily; herbal; floral; fruity
Strength	low (rank 1)
Substantivity	322 hours (neat)
Shelf life	24 months
Occurrence	cabreuva oil 57.28-60.66%; cananga oil China 20.45%; carnation absolute 14.60%; Peru balsam and Tolu balsam
Suppliers	84
Regulatory	JECFA 24, FEMA 2138, CoE 262, FLAVIS 09.727
Cosmetic uses	antimicrobial agents; perfuming agents; solvents
Oral toxicity	rat LD50 1700-2800 mg/kg; mouse 1400-1568
Dermal toxicity	rabbit LD50 4000 mg/kg
GHS	H302 harmful if swallowed; H411 toxic to aquatic life with long lasting effects
Note	blends well with alcohols and aldehydes; stabiliser for higher alcohols and aldehydes; contained in Peru balsam and Tolu balsam

Two roles

The category field reads flavor and fragrance agents, carrier solvents.

Across the hundred-odd materials I have written up, most carry one category. Few carry both fragrance material and carrier solvent, and this is the best-supplied of those (84 suppliers).

The cosmetic uses field confirms it: antimicrobial agents; perfuming agents; solvents.

The double role has a practical consequence: using it as a solvent means adding something with a smell. The strength field says low at rank 1, so the smell is small, but it is not zero.

(Of the three carrier solvents I have written up, dipropylene glycol and isopropyl myristate both read "none" for odour strength. Only this one reads "low".)

Melting point 18 to 21 C

The most important number to know before buying it.

An unheated room, a bathroom cabinet in winter, a glovebox in a cold car can all sit below 21 C. It goes solid or semi-solid in the bottle.

The appearance field says as much: "colorless to pale yellow clear oily liquid to solid".

In the clogged sprayer article I compared four co-solvents across three axes. This one has a logP of 4.00, closest of the four to most aroma materials (far better than TEC's 0.10 or DPG's -0.60), and a vapour pressure of 0.000250 mmHg, low enough. The only axis it loses on is this melting point.

Specific gravity 1.113 to 1.121

It is denser than water.

Most aroma materials sit between 0.85 and 1.05, and few exceed 1.1. In practice that means two things: the same volume weighs more than you expect, and it sinks in any aqueous system.

(One reason to write formulas in percent by weight rather than in drops, as in Why formulas should not use drops.)

The solubility field carries measured figures

For most materials that field lists only solvent names. This one gives three numbers:

musk ambrette, 1kg dissolves 450 g. musk ambrette; musk ketone, 1kg dissolves 205 g.; musk zylol, 1kg dissolves 280 g.

One kilogram of benzyl benzoate dissolves:

• 450 g of musk ambrette
• 280 g of musk xylol
• 205 g of musk ketone

The same field carries a direct statement: solvent for the nitro musks.

Nitro musks are solids (musk ketone melts at 135 C) and have to be dissolved before they can go into a formula, and this field tells you how much a kilogram will take. A number you can put straight into a formula calculation is rare in this field.

(Why measured solubility figures are so rare in that field is measured in What does 1:1.5 in 50% alcohol mean.)

That 57% in the occurrence field

cabreuva oil @ 57.28-60.66%.

In cabreuva (Myrocarpus species) oil it is more than half. The next highest are cananga oil from China at 20.45% and carnation absolute at 14.60%.

The notes field also places it in Peru balsam and Tolu balsam. Those two are regulars in the contact allergy literature, and Peru balsam sits at the top of European patch-test positive rates. That figure is in Synthetics give me headaches so I only use naturals.

It is a natural, and it makes up a substantial share of a natural balsam with a high positive rate. Those two facts do not conflict, but they are worth seeing together.

Its toxicity is a band above the other carrier solvents

Carrier solvent	Rat oral LD50
isopropyl myristate	> 16,000 mg/kg
dipropylene glycol	14,850
triethyl citrate	5,900
benzyl benzoate	1,700-2,800

Its GHS also carries H302 harmful if swallowed and H411 toxic to aquatic life, which the other three do not.

That does not make it unsafe to use; it puts it in a different band from the other three. (The notes field also mentions its use in a carpet acaricide, which is a separate direction.)

What this doesn't establish

• I have not smelled it. This is database fields plus one paper.
• "One of the few carrying both categories" is a comparison across the hundred-odd materials I have written up, not a library-wide scan. I got caught on an "only one" claim last time, so the scope is stated up front here.
• The three musk solubility figures carry no temperature. At what temperature 1 kg takes 450 g, the field does not say.
• I have not tested the practical effect of an 18-21 C melting point. "Sets in winter" is inferred from the number.
• The 57-61% for cabreuva oil is a range of literature values, not repeat measurements on one lot. How reliable that field is was measured in How much does a natural actually vary.
• I did not look into the relationship between Peru balsam's positive rate and this material. I know it is one of that balsam's constituents; I do not know whether it is the sensitising one.
• The toxicity figures are old (1948 to 1973) and are animal acute toxicity.
• What I cite Millard 2002 for is the cosolvency model; that paper did not measure this material.
• I did not recompute corpus statistics. Its 43 appearances in 954 formulas (4.5%) at a median dose of 10.5% come from the solvents article, and that corpus is of concentrates.