Conservation DistList Archives [Date] [Subject] [Author] [SEARCH]

Subject: Adhesive for glass

Adhesive for glass

From: Robert K. MacDowell <macdowell_r>
Date: Tuesday, September 1, 1998
Sincere thanks to all who took the time to respond to my initial
inquiry regarding a good choice of adhesive for consolidating cracks
in glass. Especially, I want to thank Ms. Katharina Schmidt-Ott for
sending by Snail Mail a report covering adhesive tests done by her
institution in Switzerland.

Concurrent with the flurry of information exchanges, we were
evaluating an adhesive that had not been mentioned, and, as I have
sent this response to Ms.Schmidt-Ott, I thought it might be helpful
to offer the text for general comment. A copy follows:

    Dear Ms. Schmidt-Ott,


    I have been doing some experimental work with a Loctite product,
    No. 302, Part Number 28422. This is glass-clear, has a very low
    viscosity, wicks into cracks very well, and cures by exposure to
    UV or visible light. This morning, just for fresh data, I tacked
    together two pieces of window glass about 2-cm. square,
    overlapping the edges by about 0.5-cm. The initial tack was done
    with a hand-held UV/visible unit made by Teklite. I then left
    the pieces out in sunlight for about 30-minutes, after which I
    could not break the bond by very aggressive hand pressure. The
    resultant bond was glass-clear. Then, to check for
    reversibility, I heated the pieces, starting at 250 deg.F., then
    275 deg., and finally 300 deg., at which point I was able to
    separate the pieces by hand. The adhesive remained glass-clear,
    so one might assume that elevated temperatures are not likely to
    cause discoloration.

    The first trials of this adhesive were done here in July, in
    which case I cut a piece of window glass, using a glass cutter,
    then laid the pieces flat, aligned the broken edges, and wicked
    adhesive into the crack. This was cured with the Teklite unit
    and sunlight. I did not want to break this joint, but it held up
    to fairly aggressive twisting, by hand. The purpose of this test
    was to determine if the adhesive had useable strength (which it
    obviously does) and to keep the sample to determine if any
    discoloration would take place. Not much time has elapsed, but
    the sample remains glass-clear.

    I have not pushed Loctite re the exact formula for their
    adhesives, but they are acrylic or methacrylate. This one does
    not contain any chemistry to respond to primers or accelerators,
    so the risk of discoloration is minimized.

    We all know, of course, that reversibility is an issue to which
    we must be sensitive, however there will be instances where a
    procedure to stabilize cracks is deemed necessary, and the
    adhesive we cite is probably more reversible and less likely to
    discolor or cause damage than any other that we are aware of.

    This morning I spoke to Loctite , and they gave me their Web
    address: <URL:> Also, the telephone number
    of Loctite in Germany: 49-89-92680. Fax Number is

    Incidentally, for background, you should know that we have been
    engaged exclusively in this profession since 1970, and have
    continued over the years to seek better methods and materials
    with which to accomplish various tasks in our specialty fields
    of ceramics and glass. We have done extensive projects for major
    institutions and individual clients, worldwide, and can report
    that we have never received any complaint or notice suggesting
    that any product made by Loctite has failed to perform
    satisfactorily. We began using Loctite products in 1973.

    Please let us know if you decide to do your own testing on this
    particular adhesive, as we very much value any input we receive.

    Robert K. MacDowell

                  Conservation DistList Instance 12:23
                Distributed: Tuesday, September 1, 1998
                       Message Id: cdl-12-23-004
Received on Tuesday, 1 September, 1998

[Search all CoOL documents]

URL: http://